POPULATION

about two-thirds of the inhabit- ants of the world are now periodically enumerated by means of a census. whenever a census is taken questions are asked re- garding some or all of the following matters: age, sex, occupa- tion, place of residence, religion and others. thus, regarding two-thirds of the inhabitants of the world there is a considerable volume of information of a fair degree of accuracy. the ques- tions asked vary considerably from country to country. since information is nearly alway: required regarding age, sex and marital condition, the inhabitants of different countries, the popwation of which is periodically enumerated, can be compared in respect to these matters. but with regard to the asking of questions concerning other matters much variaticn exists, and comparisons are often ren- dered impossible. ‘there is no question regarding religion in the english census, for example, and comparisons in this respect between england and other countries where such information is collected are therefore not possible. for information regarding the remaining third of the world population it is necessary to rely upon estimates. no great degree of confidence can be placed upon these estimates, and a wide margin of error must be allowed for. furthermore, estimates can only be made for the total population. it is impossible to estimate the distribution of the population in respect to age or sex, and as to this, therefore, for the remaining third of the world population there is no informa- tion at all. the first step is to analyse and tabulate this information so as to throw light upon such points as it is desirable to investigate. the information contained in recent enumerations can be com- pared with that contained in former enumerations. the next step is to enquire how the population of any area comes to be what it ts and how the changes found to have taken place bet ween one enumeration and another have been brought about. this involves an examination of the facts regarding marriages, births, deaths and migration which are generally recorded annually for those areas in which a census is taken. from these registration statistics it 1s possible, for example, to throw light upon the pro- portion between the sexes as disclosed by the census figures. it can be shown that there is a certain proportion of male to female births and that the death-rate differs for males and females in each age period. in this manner an explanation is provided of the facts given by the enumeration. ‘the attempt can then be made to push the explanation further back. the problem as to the cause of the difference in the proportion of male to fe- male births can for instance be investigated. but such enquiries pass beyond the field of what is usually included in the study of the population problem, and attention in this article is con- fined to the study of the facts as disclosed by censuses and esti- mates and to the immediate explanations of these facts as far as registration statistics throw light upon them. 1. world population and its distribution geographical distribution. of the many estimates of the pop- ulation of the world that have been made the best known are those of mm. levasseur and bodio for the years 1880 and 1900 and of m. bunle for 1920. since the census figures and estimates used 186 in making these calculations were those available nearest the years 1880, 1900 and 1920 the totals given below should be read as estimates of the population ‘‘ about ” 1880, 1g00 and 1920 respectively. thus the british figures included in the 1920 esti- mate are those given in the census of 1921. table i. estimated area and population of the world population in area in ee : mitlions in thousands of square miles density per square mile in 1920 region es ee, europe . 3,899 347 401 451 116-5 africa 11,766 197 135 132 11-0 asia 16,173 789 850 | 956 59-6 n. america . 9,891! 80 107 144 14-5 s. america 7,096 2 43 67 9°3 oceania 4,231 38 52 69 16:3 totals 53,056 [483 .. 1 [including approximately 3,222,000 sq. m. of arctic and sub-arctic territories. in compiling these estimates the malay islands have been included in oceania, whereas in other estimates they are some- times included in asia. to what extent can these estimates be relicd upon? no very high degree of accuracy can be attributed to the totals of the earlier estimates. as m. bunle points out. while the areas of the continents given in the above table may be taken as approximately accurate, the 1920 figures for africa and asia and to a lesser extent for the other continents, with the exception of europe, differ from those given by mm. levas- seur and bodio when making the earlicr estimates. the areas have of course not in fact changed, and the differences are duc to more accurate recent measurements. it is evident that if the areas were not even known with accuracy when the earlier esti- mates of population were made, these latter estimates are likely to have exhibited a much larger margin of error. that the esti- mated population given in the above table for 1920, however, is not likely to be seriously in error may be deduced from the fact that it agrees fairly closely with other independent estimates. thus the international institute of agriculture at rome gives the estimated population of the world as 1,820 millions for 1921 and the institut international de statistique 1,791 millions for 1920 and 1,895 millions for 1924. the various continents —considering the continents in turn, the figures for europe and north america may be accepted as approximately accurate, though some doubt arises concerning the figures for russia and mexico. it is thus seen that whereas the population of europe has increased by about 50,000,000 in each of the periods 1880-1900 and 1900-20, the population of north america has increased by about 25,000,000 in the first period and by about 37,000,000 in the second period. the differ- ence between the two continents in this respect in the later period is partly to be explained by migration from the former to the laiter and partly by the effects of the world war. the figures for oceania are probably more accurate than those given for any of the other remaining areas. in twenty years the population appears to have increased about 30%, which is approximately the rate of growth found for the population of north america. while very considerable progress has been made during recent years in the enumeration of the inhabitants of certain south american countries, especially of brazil, argentina, chili and uruguay, there remains a large measure of doubt as to the present population of the continent. the figures for the earlier years are still more open to suspicion. there is no question but that the population of this continent has increased largely since 1906; it may be doubted, however, whether it has increased to the extent which these figures show. the figures for asia are still less trustworthy. for some areas of that continent, india and japan for example, there are reliable figures. but for other areas, china in particular, the estimates vary widely. m. bunle accepts the estimate of 445,000,000 for china, manchuria and the chinese dependencies made in con- nection with the chinese postal administration for 1921. this estimate is among the highest, but it agrees closely with those of population the chinese maritime customs for 1922 and of the china con- tinuation committee for 1918. there are estimates as low as the 318,000,000 given by current books of reference. these low estimates have apparently been influenced by the census of 1910, which was a census of households and not of persons. further- more, the households were multiplied by different factors in different parts of the country to get the total number of persons —an obviously unsatisfactory method. given this uncertainty reserme se poe weta “j o 7) *] 7) wt | || eo ee | "i l ‘ millions. on oo i le sherry 30 20 1910 1911 1912 1913 1914 i9is t916 1917 1918 i919 8920 1921 1922 1923 1924 year. fic. 1.—chart showing the changes in the populations of certain selected countries from 1910 to 1924. the most marked features are the effect of the war upon the populations of the united kingdom (which includes the area now called the irish free state), germany, france and italy; and the rapid increase in the populations of the united states and japan. regarding the population of china, it is evident, especially in view of the tendency to accept a higher estimate, that no reli- ance whatever can be placed upon the apparent growth in the population of asia shown in the table as taking place between 1900 and 1920. the difference between the estimate for the population of asia for these dates is less than the difference be- tween the estimates of the population of china at the present day. similarly, no confidence can be placed upon the apparent decrease in the population of africa between 1880 and 1920. esti- mates for the population of the belgian congo vary from 8,000,000 to 30,000,000. in large part the difference between the figures is due to the acceptance of lower estimates for the negro popu- lation. at the same time, parts of the continent have been ravaged by sleeping sickness and other diseases in late years and it is at jeast unlikely that there has been any increase in the population. while it is not possible to obtain any accurate measurement of the proportion of the world population represented by the populations of the different continents at various dates and thus population to record the relative advance of the populations of some con- tinents compared with others, it is possible to obtain a fairly accurate measurement of the changing proportion of the total population of europe shown by various european countries. a similar calculation can be made for north and south america, though it would be misleading to take the figures back to so early a date as in europe, since it is only of late years that the data have attained a reasonable degree of accuracy. the fol- lowing table therefore shows the proportion which the popula- tion of the principal countries of europe contributed to the total population in 1880, 1910 and 1920, and the proportion contrib- uted by the principal countries of north and south america in 1910 and 1920. table ii. the proportion of the total population of europe (=100) and of america (=100) contributed by certain european and american countries at various dates north and south america country country england & wales scotland ; ireland france germany austria belgium hungary italy russia all other countries 1910} 1920 argentina brazil canada chili colombia cuba mexico peru , ; united states |s venezucla all other countries viggs i-12 1:95 11:20 13°54/14°52/13°24 6:63] 6-38] 1-42 1°65] 1:66] 1°65 4°71} 4°67] 1°77 8-52) 7°75| 8°59 25°82 |29-13/22'50 1749/17 03/3172 between 1880 and roro there were no changes in the frontiers of european countries, and the alterations in the relative posi- tions of the various countries are due to varying rates of increase. the notable features are the decline in the position of france and the advance in the position of russia. between roro and 1920 there were large losses and large gains of territory in many countries. in addition, all european countries experienced an increased death-rate. thus, in the case of the 1920 figures the changes duc to varying rates of increase are overshadowed by the effects of the war. these effects have been remarkable. it is noticeable that france, in spite of an increase in territory, has not quite maintained her position, and that germany, in spite of a large decrease in territory, has suffered comparatively little. austria and hungary now contribute less than 2% each to the population of europe. the figures for north and south america are interesting in that they show that the various coun- tries tend to maintain their relative position. | racial distribution.—it is necessary to exercise the greates caution in the use of the term “ race.”’ if by racial groups are meant groups of men exhibiting in common certain inherited character differences, then we may legitimately employ the term. there are groups of men exhibiting in common certain inherited characteristics in respect of hair, pigmentation and other physi- cal features, and it may be asked what proportion of the popu- lation of the world falls into each of the larger groups. to this question no satisfactory answer can be returned. not only in south america, for example, is the total population not accu- rately known but trustworthy data as to the proportion which each racial group, caucasian, negroid and amerindian, contrib- utes respectively to the estimated total are wholly lacking. the same applies to asia and africa. anestimate has been made by sir leo chiozza money of the total numbers of the so-called ‘“‘ white race,”’ an element of the caucasian group formerly more or less confined to europe. he finds that out of an estimated world population of 1,852 millions in 1921, there were 603,000,000 belonging to this white element, that is to say, rather less than a third of the total world population. he also finds that of the 603,000,000 about 117,000,000, or less than a fifth, are of british stock. if these 600,000,000 whites are subtracted from the total world population, leaving some 1,250 millions to be distributed among the remaining racial groups, then it may be very roughly estimated that about half this number belong to the mongoloid 187 racial group, while two-thirds of the balance belong to the non- white branch of the caucasian group and one-third to the ne- groid racial group. it thus appears that the white branch of the caucasian racial group and the mongoloid group between them account for about two-thirds of the population of the world. in the absence of accurate information as to the constitution of the population of the world to-day it is clearly impossible to discover whether the different racial groups are maintaining their relative positions or whether one or more groups have improved their relative posi- tions since 1900 or 1880. jt can only be said that the white ele- ment of the caucasian group, which was during the greater part of the last century considerably improving its position, is in any case not now doing so in the same ratio if indeed it is improv- ing it at all. it. economic distribution density—in table i. is shown the density per sq. m. of the population of each continent. the average density of population for the whole land area included in the table is 34 per sq. mile. from this table the uninhabited continent of antarctica is ex- cluded. it should be noted that no deduction has been made on account of the barren and almost uninhabitable areas which are included within every other continent. the inclusion of these areas tells least against europe and most against north america. thus greenland, alaska, the northwest territories of canada, labrador and the arctic islands, cover an area of approximately 3,222,000 sq. m., and contain a population amounting only to about 113,000. if these areas are excluded the average density of population in the rest of north america rises from about 14 to about 22 per sq. m. and in the world as a whole to about 36. taking the figures as they are, it appears that europe and asia are the most crowded continents, europe being about twice as densely peopled as asia. the remaining continents are, relatively speaking, empty. among countries and areas which are regarded as distinct units from the point of view of political organisation, barbados is the most densely populated, having (in 1921) 952 per_sq. m., while java comes next with (in 1920) 689 per sq. mile. belgium is the most densely populated among european countries with 635 per sq. m. (in 1920), while great britain has 482 per sq. m. {in 1921).. it is worthy of notice that england and wales have a density of 649 per sq. m. (in 1921) and are therefore more densely peopled than belgium. such comparisons can be widely extended and up to a point are informative. jt should be remembered that they may also be very misleading. the area of egypt is, for example, 383,000 sq. m., and the total population amounts to 12,750,000, giving an average density of 33 per sq. mile. but the population of egypt is almost wholly confined to the valley and delta of the nile, the areas of which contain 12,023 sq. m., giving a density of 1,06r per sq. m. of the inhabited area. so, too, the figure of 1-83 per sq. m. for the continent of australia misrepresents the facts. a con- siderable proportion of the continent is uninhabitable, and if this eae was excluded from the calculation the density per sq. m. would e appreciably raised. the figures for density given above fail to bring out certain important facts. an examination of the map of the world show- ing density of population without regard to political boundaries makes it at once apparent that there are three densely populated regions of no great size. the first region is that part of europe which lies south of the parallel 60° north latitude. the second ts that made up of the greater part of india and ceylon; the third includes japan, the greater part of manchuria and china and a strip along the coast of tonking. it may be roughly esti- mated that the population of these regions amounts to 440,000,- 000, 480,000,000 and 300,000,000 and the areas to 2,900,000, 2,000,000 and 1,200,000 sq. m. respectively. we thus find over 1,200 million people living on about 6,000,000 sq. m., or in other words about two-thirds of the world population living on about one-eighth of the habitable area. there are four other very much smaller regions which are densely peopled, the new england area of the united states, the nile valley and delta, the tip of the malay peninsula and java. urbanisation.—as seen above, figures for average density may be misleading if within the area of a country or region there 188 are large barren tracts. but even if they are not misleading, or if calculations are made to correct the misleading impression otherwise given, such figures fail in themselves to provide an adequate picture of the conditions under which the inhabitants of an area are living unless they are supplemented by facts throwing light upon the degree of urbanisation. it is not an easy matter to obtain a measure of urbanisation. the census returns of most countries classify the population into urban and rural, but while these figures are useful for com- paring changes in the same country between two dates, they are of little use when comparing two countries. this is so be- cause the classification is based upon very different methods in different countrics. thus, owing to differences in local gov- ernment between england and scotland, the figures giving the percentages of urban dwellers out of the total populations are not comparable. again, those officially classed as urban dwellers often live under what to most people would appear as rural con- ditions. english urban dwellers include, for instance, according to the official classification, those who live in the areas of urban district councils and, as is well known, to live in such an area is often regarded as having escaped from a town. the best method is perhaps to note, in the table below, the percentage of the total population living in towns having a population of 100,000 and over. table iii. density of population and urbanisation: 1020 | peaaeny of | . yulation lhyv- population pe 7 : | country per eas bane eae aie inhabitants and over egypt (cultivated area only) 1,061 9:9 england and wales ; 649 38-8 apan , . ' ; ; 484 12-1 taly . : : ; ; 32 13-7 germany : : 319 23-7 china . 274 2-2 france 184 15-0 india : , : : 177 2-6 russia (in europe) .. 57 4-0 united states. ; 35 25°7 canada (excluding north west territories) 3 18-9 australia 1-83 41°3 this table must be interpreted with caution. the method of calculating urbanisation here adopted is not perfect. the ur- banisation of australia, for example, is exaggerated. the urban table iv. proportion of the sexes living and at birth about iqio and 1920, in certain european countries no. of male births to 1,000 female births about 1920 no. of females to 1,000 males country born born living about about germany austria . belgium bulgaria denmark spain france . england scotland ireland greece . italy norway holland portugal rumania russia serbia sweden : switzerland . population population of australia is concentrated in five large towns and there are few towns of relatively small size. if 50,000 instead of 100,000 was taken as the lower limit, australia would come out as relatively less urbanised than under the present method of calculating urbanisation. allowing for its defects, the table does show that there is little association between density of popula- tion and urbanisation. densely populated countries may or may not have a high degree of urbanisation; australia is an example of a sparsely peopled country with a high degrce of urbanisation. t'urther examination of the table shows that urbanisation goes with industrialisation rather than with density. in other words, a high degree of density without industrialisation does not lead to urbanisation, whereas urbanisation can go a long way where industrialisation prevails, even though the population is sparse. the countries which have a high degree of density with- out urbanisation are those such as china, india and egypt, where intensive agriculture is practised. density thus accompanies intensive agriculture and may also be attained under industrial condition, there were, in 1920, 16 towns with a population of over 1,000,- e090 apiece. of them six were in europe, five in asia, three in north america and two in south america. one remarkable result of the war has been the decline in the population of certain great towns. thus between 1910 and 1920 moscow decreased from 1,481 to 1,028, leningrad from 1,907 to 706 and vienna from 2,030 to 1,841 millions. ill. distribution by sex and age sex.—it is a familiar fact that in great britain and most euro- pean countries females outnumber males. the extent to which this is the case is shown in tables iv. and [v.-a. the tables show the number of females to 1,000 males in certain european countries about the years 1910 and 1920 and in certain non- european countries about the year 1920. the european countries include all the larger ones, among them russia, germany and france, and most of the smaller ones. the newer countries e.g., czechoslovakia and yugoslavia are not included, as in their case there is no information about conditions in toro. wales, as usual in such cases, is treated with england, but separate figures are given for scotland. the non-european countries are the united states, canada, austra- lia, india, japan, south africa and egypt, for all of which the necessary statistics are available. as regards the figures for the european countries for which there is information at both dates, it is seen that in the case of the countries neutral during the war there has been little change on the whole in the proportion between the sexes, whereas in the case of belligerent countries the former considerable excess of females was increased. belligerent countries are therefore ab- normal in respect of the existing large excess of females, and the increased disparity between the sexes is due to deaths of males during the war. the exceptional case of the lately belligerent countries may be thus left out of account, and the proportions about 1910 taken as more normal for european conditions. it next appears that most european countries are strongly contrasted with those extra-european countries for which figures are given. the latter show a deficiency of females. table iv.-a. proportion of the sexes living and at birth an certatn non-european countries about 1920 no. of no. of male births to females 1,000 female births . h h ” country pee ia | about 1920 about born born 1920 living dead u.s.a. ; 960 - - canada ; 3 : : 940 1065 1402 ebyde ec 3m 8 3] oof 1093 1512 south africat 2. «| 943 1076 be vapatics a ¢- 2 ; 979 1045 1176 india . : : 928 2 si | australia. : . : 968 1062 i 1excluding coloured population. population table v. death-rate among males expressed as a percentage of the death-rate among females at different age pertods age periods country ies | aaa i-4 | 5-9 | 1o-14 | 15-19 | 20-24 | 25-34 | 35-44 | 45-54 | 55-64 england and wales 1911-5 iig 103 95 “t15 eee 125 125 128 u.s.a. registration area iqii-5 . lt. , | 120 ii4 “t10. “ts “118, 116 a2 127 age periods country under i i-19 20-39 japan 1918 125 | 88 81 | the next step is to examine the sex ratio at birth. the differences noted might be due to an excess of female births in some countries and of male births in others. the figures in the table show that this is not so. for all countries for which figures are given there is an excess of living male births, accepting this as a fact, how, it may be asked, is the excess of females in europe and the deficiency of females in other countries arrived at? table v. shows that in eng- land and wales and in the united states registration area during the years 1911-5 females in every age group experienced a lower death- rate than males with the one exception of the age group 10-15 in england and wales, where the males have the lower death-rate. it thus comes about that, although in the earliest age groups males outnumber females because more boys are born than girls, females soon come to outnumber males and in fact preponderate in the age group 15-20 in england and wales and in all succeeding age groups. thus, taking the population as a whole, females outnumber males. it may then be asked why females have this more favourable expe- rience. there are two reasons which account for this fact. the fe- male sex is more hardy than the male sex, and on the whole resists disease and the wear and tear of daily existence better than does the other sex. again, females are less exposed to accidents than males. thus, industrial accidents, as exemplified by mining disasters and explosions, fall more heavily upon the latter than upon the former. men, in fact, lead a more hazardous existence, and more often lose their lives at work or at play than do females. the out- is of war is only a special case of the extra exposure of males to risk. there are, however, countries where males predominate. this is so in the united states, for instance, and since table v. shows that the male death-rate there exceeds the female death-rate to nearly the same extent that it does in england some explanation is clearly required to account for the difference between the united states and england in this respect. the explanation is to be found in the fact that large numbers of emigrants from europe have entered the united states during recent years, and that among these emigrants males preponderate. this tendency for males to migrate more than females accounts also in some degree for the deficiency of males in those european countries from which movement takes place. the excess of males in the united states is found chiefly among the foreign- born whites and not amongst the native-born whites and negrocs. again, in the denscly settled regions of that country females outnum- ber males. the same factor explains the excess of males in the british dominions. this factor cannot, however, be invoked to explain the position in india, japan and egypt. table v. shows that in japan the female sex experiences a higher mortality than the male sex between the ages of 1 and 40. this higher mortality is clearly the cause of the excess of males in that country. it is outside the scope of this article to discuss the determination of sex. the figuresin table iv., however, carry us somewhat further back. so far, reference has only been made to the figures for living births, which show that more males are born alive than females. reference may now be made to the figures for still births, and it will be observed that males bear a higher proportion to females among still births than they do among living births in every country for which figures are available. males thus suffer a high mortality at birth, and the original disproportion between the sexes must be higher than the statistics for living births show. evidence derived from table vi. observations upon miscarriages and from other sources goes to show that males suffer a disproportionately heavy mortality at a still earlier stage during pre-natal fife. it thus follows that the propor- tion of males to females must be very much higher at conception than it is at birth. it may be observed that modern investigations into the determination of sex have shown that sex is determined at conception and cannot be influenced by any method now known. nor is any method of so doing likely to be discovered. distribution by age.—it is in theory possible to have a popu- lation distributed in almost any fashion according to age. thus, an uninhabited island might be peopled by boys or girls or by persons 60 years of age and over. when the populations of different countries are examined it is found that there are re- markable differences in respect to age distribution. thus, one country may have a relatively large number of young persons and another a relatively large number of old persons. these differences are of interest from two points of view. first, the proportion of persons of working age is important because upon their efforts depend the rest of the non-working population, both young and old. again, only when the age dis- tribution is known can the figures for birth-rates and death-rates be interpreted. it may be known that there are a certain num- ber of births per 1,000 of the population and a certain number of deaths per 1,000 of the population in a certain area during a certain year, but, unless it is also known how many wives of child-bearing age and how many old persons there are in every 1,000 of the population, this information is of little use in ena- bling comparisons to be undertaken between one country and another or between one town and another. thus, old people might retire in large numbers to a healthy town and live there until their death, thus causing a high death-rate and giving the appearance of bad conditions in what was in fact a healthy town. table vi. gives some information as to the age distribution in certain countries:— figures such as these are not of a very high order of accuracy. ignorance and mis-statement vitiate the data to some extent. there seems to be a tendency for persons below a certain age to un- derstate their age and for persons above a certain age to overstate their age. this may be in general attributed to vanity, the young being proud of their youth and wishing to appear even younger and the aged being proud of their advanced years and wishing to appear even older. nevertheless, the figures are of sufficient accuracy to bring out some remarkable differences in age distribution as between different countries. the influence of the decline of the birth-rate is best seen in the case of england and scotland when the age distribution of ig1i is com- pared with that of 1921. the number of those under five years of age has decreased, while there has been a marked increase in the number of those over 45. if the number of those of 65 years and over number of persons at different ages per 1,000 of the total population about 1911 and 1920 eng- england rend ; , . ger- | prus- : aes ea ies scotland usa australia france many sik japan he iqii 1920 iqoie 1920 igtt 3 1920 iqii 1920 iqt1 1911 1920 1920 1881 under § . 107 88 112 97 116 110 119 116 89 120 74 129 136 5-15 199 189 210 198 205 209 199 206 169 220 221 222 229 15-25 182 176 185 | 186 197 177 202 167 160 183 194 179 188 25-45 299 293 282 | 276 292 2y6 288 298 291 273 284 261 259 45 and over 213, 254 211 | 243 190 | 208 ig2 213 | 291 204 227 209 _ 188 all age 1,000 | 1,000 | 1,000 , i,000 | 1,000 | 1,000 | 1,000 ' 1,000 | 1,000 | 1,000 } 1,000 | 1,000 | 1,000 190 is considered, it is found that there were in every thousand of the population 52 in this age group in ig11 and 60 in 1921. since persons of these ages are for the most part dependent upon others for their support, it will be seen how the burden increases which a declining birth-rate places upon the producing section of the community. there is a close similarity between the figures for the united states and australia. both countries have experienced the influence of immigration and of a declining birth-rate. the figures for france and germany are pre-war figures. the birth-rate of france has long been steady, and it is towards some such age distribution as that now seen in france that the age distribution in england is tending. figures for the whole of germany since the war are not available but, in so far as prussia is representative of german conditions, the influence of the decline of the birth-rate upon the age distribution can also be seen in germany. in the last two lines of the table the figures for japan in 1918 and for england in 1881 have been set out, ancl it is evident that the present constitution of the japanese popula- tion in respect to age closely resembles that of the population of england and wales 40 years ago. how do these differences arise? if the birth-rate and the death- rate remained constant in a country over a period of years and if there was no migration, the population would become and remain distributed according to age in a certain fashion. it is because the birth-rate, death-rate and migration rate change in the same coun- try as between one period and another and because they do not change similarly in different countries that there are differences in age distribution as between the population of the same country at clifferent times and between different countries at the same time. the case of england and wales may be taken as an example. during the last century the death-rate declined, and principally as a result of this the population rapidly increased. the population of england and wales in 1861 was 20,066,224, and in 1921 was 37,886,699, and in consequence persons of 60 years of age and over are the remnant of a population little more than half as large as the existing papulation. table vii. 1876-85 1886-95 country ——. m f, m. f england and wales... 25:9 24°4 26:4 249 france. ; 28-0 239 27'9 23°4 germany . . ii a a holland . : 28-3 26°5 28-0 26:1 italy . ; o re : scotland oe - - es sweden 28:7 271 28-7 27:0 switzerland a ug 28-5 26'1 but the birth-rate has been decreasing for half a century. it can be calculated that, if the birth-rate remains at about its present level and if the death-rate does not change significantly, the population will cease to increase within two decades, and will remain approxi- mately stable. under these circumstances the age distribution would become stabilised. this example explains the differences between the age distribution in france and england, since in the former country there has been little change for some years, with the result that there is a larger proportion of old persons than in this country. migration largely explains the high proportion of younger persons in the dominions, and similarly explains the deficiency of such persons in ireland. iv. marriage, birth and death rates marriage —the age distribution of the population is given in the census returns, but the attempt to interpret changes in the age distribution leads to the employment of data regarding birth- rates and death-rates which are obtained from registration sta- tistics. similarly, certain data regarding marriage can be ob- tained from the census returns, but for the interpretation of these figures recourse must be had to registration statistics of marriage. one method of representing the facts is to give the number of married persons per 1,000 of the population. this method is open to the objection that, since the age distribution in the same country at different times and in different countries at the same time is not the same, figures are not comparable. another method is to give the number of married persons per 10,000 of the marriageable population. the marriageable popu- lation is defined as consisting of all males 18 years of age and over and of all females of 15 years of age and over, excluding only those of either sex who are married. this latter method is to be preferred, though it is applicable only to countries with euro- pean standards and habits. it fails to represent, for instance, the true condition of things in india, where in 1921 there were 110,- population 684 bovs and 218,463 girls married under five vears of age and 727,405 boys and 2,016,687 girls married between the ages of five and 10. further, there were over 2,000,000 boys and 6,000,000 girls married between the ages of 10 and r4. the table brings out the fact that the different countries vary in a striking manner in respect of the marriage rate thus calcu- lated. there is a group of western european countries where the rate tends to exceed 500, and a group of eastern european coun- tries where it approaches or exceeds 1,000. the rate has not tended to vary widely in each country during the period to which the figures apply. generally speaking, countries that had a high rate around 1876 have maintained that rate, and coun- tries with a low rate have maintained a low rate. it has long been known that the marriage rate in england tends to vary in the same direction as the price of wheat, the value of exports, and such indices of prosperity as the amount cleared at the bank- ers’ clearing house per head of population. but these variations are not large, being in the same direction but not in the same de- gree as the changes in the indices of national prosperity. another matter of considerable importance is the mean age at marriage. in giving figures for the mean age at marriage it is best to omit marriages of widowed and divorced persons, it is the age at marriage of bachelors and spinsters that is of importance, and the facts as to the age at marriage of these persons are obscured if other marriages are taken into account. the mean age at marriage is of course not the same as the most usual age at marriage. thus, the mean age at marriage in england and wales in 1920 was not far from 28 for bachelors and 26 for spinsters, whereas the most common ages at marriage in the same year were 24 for men and 21 for women, mean age at marriage of those not previously married 1896-1905 1906-10 iqii-i5 m. f nl. f m. f. 26°8 25-3 27-2 25:6 27°4 25°7 27°9 23-6 28-0 23°7 eae — os - re a 27-4 24°7 27°9 25:9 27°7 25°9 27°6 25°8 27-4 23:9 27°1 23-6 27°2 23-6 27°6 25°6 27°8 25°8 27°8 25°8 28:6 26°6 28°7 26-3 28:8 26-4 28-4 26-0 28-3 25:8 = the table shows that the mean age at marriage is very similar in all countries for which figures are given, the age for men is a little over 27, while that for women varics from 23-26. furthermore, there has been but little change in the mean age for fifty years: in france it has remained practically the same; in england and wales there has been a tendency towards postponement; whereas in holland the tendency has been the other way. births, deaths and natural increase —the census provides no information as to births and deaths. weare here dependent upon registration statistics. the most usual method of recording the birth-rate and death-rate is to give the number of births and the number of deaths per 1,000 of the population in any year. for certain purposes this method is open to objection. unless the age distribution is known and allowed for, it is clear that these crude rates are no measure of the true fertility of married women or of the health and vitality of the population. but these crude rates serve a useful purpose. the difference between the rates represents the natural increase or decrease per 1,000 of the popu- lation as the case may be. this natural increase or decrease is an important fact in gauging the actual position of any country, though further analysis is required in order to throw hight upon the manner in which the natural increase or decrease is brought about. the crude birth-rates varied from 40-3 in bulgaria to 19-19 in switzerland during the period under review. the crude death-rates varied from 33-6 in greenland to 9-1 in new zealand. the highest natural increase was in bulgaria and the lowest in france. while it so happens that bulgaria had the highest crude birth-rate and the highest natural increase, a glance at the table will show that there is in general no correspondence between a high birth-rate and a large natural increase. thus greenland, with a birth-rate nearly as high as bulgaria, had, next to france and austria, the lowest natural increase. the table shows that countries with a high birth-rate often have a high death-rate and thus do not exhibit a jarge natural population mean annual number of marriages per 10,000 marriageable persons table viii. country period period a austria . 1908-14 536 | 1876-85 551 belgium ae 537 | 1876-85 437 bulgaria . i9lo-1t 122 1896-1905, | i,iil denmark : : 1908-14 515 | 1885-94 540 england and wale - 501 | 1876-85 568 france . ; ‘ 540 | 1875-86 496 holland . 522 | 1875-84 536 hungary ‘ 844 | 1876-85 1,038 ireland . f ‘ ba 253 | 1876-85 260 italy. sy 5 es 578 | 1877-86 545 norway . . ; : a 416 | 1881-85 458 prussia . . , ‘3 588 | 1876-85 578 rumania . 1912-13 g90 | 1896-1903 873 scotland so 1908-i4 409 | 1876-85 445 secdia so ys 1909-12 1194 | 1896-1905 | 1,386 spal ss es 526 | 1900-1 652 sweden . ss 367 | 1876-85 417 switzerland 1878-85 increase. the association of a high birth-rate with a high death- rate has been much discussed. what the precise nature of the association may be is not clear; to some extent, no doubt, a high birth-rate renders the care of children by the mother more difficult, and to this extent a high birth-rate may be said to cause a high death-rate. at the same time, however, a high death-rate tends to ca)l forth the birth of more children to fill the vacant places, and to this extent the relation is reversed and a high death-rate is the cause of a high birth-rate. it is worthy of note that, as a comparison of table viii. with table ix. shows, those countries with a high birth-rate are in general also those with a high marriage rate. but the most interesting and important fact which emerges from table ix. is that the same natural increase is found in countries which have widely divergent birth and death-rates, thus, the united states and japan had both the same natural increase for the periods under review, but, whereas the former had relatively low birth and death-rates, the latter had high rates. the contrast is even more striking between new zealand, with its low rates and natural increase and such a country as ceylon, with a death-rate nearly as high as its high birth-rate and in consequence a jow natural increase. apart from the question whether increase is desirable and if so to what extent, it is clear that, if there is to be increase at all in any country, those conditions are to be preferred by which it is at- tained by a margin between a low birth-rate and a low death-rate rather than between a high birth-rate and a high death-rate. table ix. birth-rates, death-rates and natural increase crude birth- rate ct a | ee | natural increase country period australia i9iq-24| 24:6 10-4 14-2 austria 1919-23] 21:9 17-9 4:0 belgium 1919-24] 20°5 14-0 6°5 bulgaria 1920-1 40°3 21-7 18:6 canada 1920-4 25:3 10-9 14-4 ceylon 1919-23] 38-6 31°6 7-0 chili i9ig-24| 39°7 32-6 71 denmark . ; ‘ i9ig-24| 22:9 11-8 ii! england and wales . i9ig-24| 21:0 12°5 8:5 egypt er 1919-23} 41-6 26:6 15:0 france : : : 1920-4 20:0 17°4 2:6 germany , 1919-24] 23°5 14°8 8-7 greenland . p ; igig-22| 38-7 33:6 5-1 holland . . i9i9-24| 26:6 ii-4 15:2 hungary ‘ : i9iq-24} 29:9 20°9 9:0 italy . , ; i9qi9q-23} 27:3 17:0 10°3 japan : ; : 1919-22] 34:7 23°6 ii-! new zealand : i9i9-24| 23:0 q-i 13°9 norway. iy i919-24| 23:7 12-2 ii'5 philippines ‘ : 1921-3 32°8 19-0 13°8 rumania es i919-22| 32:9 22-9 10:0 scotland ‘ iqi9q-24 | 23:9 14°4 9°5 south africa : ig1g-24 | 27°5 10-4 17°1 spain . * ‘ee *s 1919-23] 29°8 21°8 8-0 sweden i9iq-24| 2073 12-8 7°5 switzerland : 1919-23| 19°9 13:2 6-7 lsa; a i9igq-23 3° : uruguay 1919-23 while natural increase is an important fact, it should not be taken at its face value. it should not be assumed, for example, that a given rate of natural increase will continue in any country, even if the true ioi fertility of married women and true mortality of the population re- main at their existing strengths. this is so because the crude rates are imperfect measures of fertility and mortality. the crude birth- rate is the number born in any year per 1,000 of the population. in order to get a measure of true fertility it is clearly necessary to relate the births to the wives of child-bearing age in the population. in this manner the crude birth-rate can be ‘' corrected.’ newsholme gives the following example of using this method of correction. in ireland, the crude birth-rate was 24-5 in 1881 as compared with 33-9 in england and wales, and 23-1 in r90r as compared with 28-4. when these crude rates are corrected they become for ireland 35-2 in 1881 and 36-1 in 1901, and for england and wales 34-7 in 1881 and 28-4 in 1901. thus, the true fertility in ireland was in fact as high as that in england and wales in 1881 and much higher in 1901. the explanation of the remarkable differences between the two countries lies in the fact that, whereas in england in 1901 the num- ber of females aged 15 to 45 per 1,000 of the total population was 250 and in ireland 235, the wives at ages 15 to 45 in england formed 46-8° and in ireland only 32:5 % of the females at the same age. in addition to the importance which these corrections have as throw- ing light upon the true fertility of a population, they are also relevant to the problem of natural increase. let us take the case of england and wales as an example of this fact. the figures show that there is still a considerable natural increase. it is, of course, true that, if the crude birth and death-rates remain unchanged, the population will continue to increase. lt is popularly supposed that these rates represent the true fertility and mortality of the population.. as shown above, this is not so. the true fertility and mortality of the population can be calculated, and it can be shown that, if the true fertility and mortality of the population remain as they now are, the crude rates will change until there is no gap between them. in other words, the true fertility and mortality of the population remaining as they now are, natural increase will be- come negligible in two decades and will later disappear. the explanation of this paradoxical result is as follows: the age distribution in england and wales is peculiar in that there is a relatively small number of persons in the older age groups. as a result of the decline in the birth-rate, the relative number of these older persons will increase and in consequence the death- rate will increase though the true mortality remains the same. the considerable natural increase now shown is, therefore, illu- sory, and will disappear unless there is a marked decrease in the true mortality or an increase in the true fertility. should there continue to be a decrease in the true fertility as is in progress at present, then there will be a cessation of natural increase at a still earlier date. the decline in the english birth-rate—the crude birth-rate in england and wales was 33-5 for the period 1871-5 and has declined to 18-8 in 1924. the decline has been in evidence in most european countries and in countries peopled by persons of european descent, though it usually began later than in eng- land. enough has been said above to show that these crude rates require interpretation. the decline in the crude birth-rate might be due to a decrease in the number of marriages, to a decrease in the proportion of women of child-bearing age, or to a decrease in the number of children born to married women of child-bearing age. a reference to table viii. shows that, while there are nota- ble differences between different countries as regards frequency of marriage, there has been for so years little change in this respect in different countries. in other words, the decline in the birth-rate is not explained by a decline in the frequency of marriage. . | table vii. shows that in some countries there has been a tend- ency towards postponement of marriage, notably in england, but not in others. now since the decline in the crude birth-rate is almost equally well marked in countries where there has been no postponement, the inference is that some factor or factors other than postponement of marriage must have played the prin- cipal part. but in england postponement has not been a negligi- ble factor. it has had the effect not so much of decreasing the proportion of married women who are of child-bearing age as of decreasing the proportion of married women who are of the most reproductive ages. thus of 1,000 wives under 45 there were in iqoi, 593 under 35 and in 1921, 535 under 35. nevertheless, in england, as elsewhere, the chief cause of the decline has been 192 the fewer number of children born to women of child-bearing age after making allowances for the smaller proportion among them of women of the most reproductive ages. the problem is thus shifted to an inquiry into the reason for this declining fertility of married women of reproductive ages. many suggested explanations can be shown to be inadmissible. in particular, it has been suggested that there has been a decline in reproductive capacity. now reproductive capacity is a biologi- cal characteristic, and there are no biological grounds for sus- pecting any such change. further, had any such change taken place it is only reasonable to imagine that it would have affected all sections of the population alike. but it is well known that it has not equally affected all sections, the birth-rate of the so- called upper classes having declined to a far greater extent than that of the so-called lower classes. by a process of exclusion the position is reached that the cause of the decline is to be sought in a change of habits. there is naturally but little direct evidence of such a change but there is an overwhelming amount of evi- dence that a change of habits has been advocated in the shape of the adoption of birth control methods. it is only reasonable to suppose that as a result of this propaganda birth control (q.v.) has been put into practice on a wide scale in many coun- tries and that the decline of the birth-rate finds here its chicf explanation. (a. m. c.-s.) bipliograpiy.—the census reports of the different countries; 5s. w. wynne, practical uses of vital statistics (new york, 1918); a. m. carr saunders, the population problem (oxford, 1922), and population (london, 1925); h. cox, the problem of population (london, 1922); raymond pearl, the biology of death (1922); i. s. falk, essays on vital statistics (new york, 1923), and principles of vital statistics (new york, 1923); e. b. reuter, population problems (new york, 1923); g. c. whipple, vital statistics, 2nd ed. (new york, 1923); h. wright, population (london, 1923); e. m. east, mankind at the crossroads (1923); a. newsholme, elements of vital statistics in their bearing on social and public health prob- lems (new york, 1924); raymond pearl, studies in human biology (1924); the biology of population growth (1925). port elizabeth, s. africa (see 22.112), had a population (xg11) of 37,063, and (1921) of 46,834, of whom 23,935 were whites, mostly of british descent. it is the third port of the union, a centre for the wool and ostrich feather trade, and imports general merchandise for a wide area. among modern buildings are the provincial hospital, the grey institute boys’ high school, and the grand theatre, the last in main sireet. a breakwater intended to be 8,500 ft. long, part of a scheme to provide a completely sheltered harbour, was in 1925 under construction at the dom pedro jetty. in 1913 the tonnage of ships entering the port was 2,064,000. there was a marked fall thereafter which reached its lowest point (445,000 tonnage) in 1918, by 1923 the standard of 1913 had been, however, sur- passed, the tonnage being 2,143,000. in that year the cargo handled at the port was 421,000 tons. the average value of imports, 1910-4, was £8,856,000, of exports £3,679,000. in 1923 the figures were, imports £12,519,000; exports £6,381,000. port engineering (sce harpowr, 12.935; dock, 8.353; breakwater, 4.475; jetty, 15.359; pier, 20.588; river and canal engineering).—since 1910 important developments have taken place in many parts of the world to keep pace with the advance of engineering science, and particularly with the increase in the dimensions of shipping which was so marked in the io years preceding the world war. the dimensions of shipping in 1910 the largest merchant ship afloat was the “ mauretania,” of 30,696 gross tons, 762 ft. in length between perpendiculars, and 88 ft. beam; but, excluding transatlantic liners, the largest cargo- carrying ship did not exceed 13,000 tons. in 1914 merchant ships far exceeding the dimensions of the “’ mauretania ” were either afloat or under construction. harbour authorities, not only of those ports which accommodated the great transatlantic liners but in many other parts of the world, were engaged in increasing their facilities to serve shipping of far larger dimensions than any which had hitherto used their navigable waters. during the war period harbour develop- ment throughout the world was arrested and confined largely to works of a nature essential to the objects of the belligerents. since i914 no ships of greater dimensions than those built or i port elizabeth—port engineering building at that date have been laid down and it is unlikely that there will be any marked increase in the maximum dimensions of trans- atlantic liners; the dimensions of the ‘‘ majestic ”’ (ex. ‘‘ bismarck ’’), claimed to be the largest now (1926) afloat, and completed in 1921, being §6,551 gross tons, 915% ft. length between perpendiculars, and 100-1 ft. beam. this vessel, when fully loaded, had a draught of 38-9 fect. on the other hand, a considerable increase seems likely in the number of cargo-carrying ships of large capacity. in 1910 general cargo vessels of over 10,000 gross tons were rare. ships of this class up to 6oo ft. in length, 72 ft. beam and 18,000 to 20,000 tons gross were not uncommon in 1926. depth of water.—the deepening and widening of the suez canal enabled ships of the largest class to trade with eastern ports. the canal has now been deepened throughout to 12 metres, and ships drawing 33 ft. pass through it. it isintended to provide for a draught of 35 feet. concurrently with the deepening and widening of the suez canal, the more important ports of the far east have been developed with the object of accommodating the largest ship capable of passing through it. the opening in 1914 of the panama canal, having a navigable depth of 40 ft., also had an important bearing on the development of harbours in the far east and on the pacific coasts of amcrica. the gencral result is that whereas in 1910 a navigable depth of 30 ft. was considered ample for practically all requirements, a depth of 35 ft. is regarded essential in harbours of the first class, and in special cases even greater depths are aimed at. effect of depth on cost—the cost of constructing port works in gencral shows an increase approximately proportional to the ratios of the cubes of the draughts of the largest ships for which they are designed.! the provision of port accommodation for shipping of the largest class is important from the point of view of finance and may become one of balance between the relative economy of ships of large draught and cargo capacity and the capital cost of port con- struction, | materials and equipment . materials of construction——there has been no radical change in the nature of the materials used for the construction of port works since 1910, although reinforced concrete (see ferro-concrete) has been increasingly used in building jetties, wharves and other sea works. rapid-setting aluminous cement was introduced on the con- tinent during the period under review, and its employment in great britain is extending, but its suitability for use in sea water had not been fully tested by the lapse of time. the use of portland cement concrete for all kinds of port construction, quay walls, docks, break- waters, etc., became universal, and its cheapness in comparison with the cost of dressed stone in the large majority of situations re- stricted the employment of the latter for such purposes as copings, sills, altar courses, etc. as an example of this change in practice the rosyth dockyard may be mentioned. practically all british admi- ralty docks built prior to 1910 are faced throughout, or at any rate on all important surfaces, with dressed stone, usually granite. in the case of rosyth, however, the quay and dock walls are faced with granolithic concrete, dressed granite being used only for copings, quoins, sills and in similar positions. the use of greenheart for the construction of dock gates gave place to mild stecl, largely owing to the high cost of suitable timber in comparison with steel work. mechanical equipment of docks ——the most marked change in port cnginccring, apart from the increase in dimensions, is in con- nection with the mechanical equipment of docks and the provision of storage and traflic facilities. the use of electric power for cranes and other mechanical appliances is almost universal, and even in ports where hydraulic power is available electric appliances are in- stalled side by side with hydraulic equipment. hydraulic power presents undoubted advantages for operating gate-opening ma- chinery, coal hoists and some other forms of dock equipment, but the adaptability and flexibility of electric power led to its general adoption. ‘the mechanical handling of cargoes has made enormous strides. elaborate and costly plants for storing, loading and dis- charging grain, chilled meat, coal, ores and oil in bulk have been established at ports in all parts of the world where such cargoes are dealt with in considerable quantities. the equipment of ware- houses and transit sheds with electrically operated cranes, run-ways, loaders, conveyers and hoists is general, and high-speed electric cargo cranes are provided on a large scale at most european and north american ports, although in the far east cargo is largely discharged by means of ship's derricks. floating and fixed cranes of great capacity for lifting exceptional loads are common, many floating cranes, lifting 250 tons, are in use: and one at least in america, a fixed crane at the philadelphia navy yard, was capable of lifting 350 tons. breakwaters and harrours breakwaters—in 1910 the greatest depth of water in which a breakwater structure had been founded was probably at naples, where the outer portion of the san vincenzo breakwater was constructed at a depth of about 120 f{t. below mean sea-level. the outer portions of the valparaiso breakwater,? built 1912-21, were founded at 1 see l. h. saville, xjjjih international congress on navigation (london, 1923). 2 proc. inst. of c. e., vol. 21.4 (1921-2). port engineering 2afe rin yen concrete monolith mean sea level — prabereeas: “a buren aas stone rubble /47 ft. ri aves : (ca 25 $0 ba] cas o5sd bir bse fe ; deen bz aq eds © fe quarry waste 2 sand . a, 0 -, of "att bart yr “fg ae e = han =: e, : . 7 . *. . oe « . . e * . . id weak: concrete: ban a we 6 oh is ao : tm ry - 4 =, = of * cea ae « dredged bed of basin io 5 0 10 fic. 1.—valparaiso breakwater. fic. 2,—dublin quay wall floated monolith of reinforced concrete. a depth of 147 ft. below mean sea-level, and an extension 2,300 ft. in length, was to have its termination in a depth of about 180 feet. the breakwater is of composite construction, the inner portion con- sisting of concrete blocks set on a submerged rubble stone mound, and the outer portion (fig. 1) of reinforced concrete caisson monoliths, which were floated over and sunk upon a rubble mound (with sand core), the surface of which is at a level of 46 ft. below mean sea-level. these monoliths are the largest which had been hitherto constructed in such a situation, each measuring 20 metres long by 16 metres wide and i5 metres in height, and were, after sinking, filled in with mass concrete. the monoliths are surmounted by a mass concrete superstructure carried to a height of 7-5 metres above mean sea-level. the use of concrete monoliths in breakwater and pier construction has become gencral. reinforced concrete caissons, floated into position and filled in after sinking with mass concrete, have been utilised both for the construction of super- structures resting on rubble mounds and, less frequently, as mono- liths resting on the sea bed. other examples of the former are the galleira mole at genoa and the extensions of the breakwater at naples. the latter system was employed in the construction of some of the heligoland moles demolished after the war. the breakwater at fishguard, constructed by the great western railway co. between 1900 and 1918, is an interesting example of a composite breakwater of modern construction which, in spite of weaknesses inherent in the original design, has been made suitable to the exposed situation in which it is placed. originally designed as a rubble mound carried up to about the height of high water with concrete superstructure and parapet, the wastage of rubble on its seca face was found to be so serious that in the year 1913 it was de- cided to protect the rubble by the deposit of 40-ton conercte blocks placed “‘ pell mell ’’ over the whole of the seaward face of the mound above the level of l.w.o.s.t. previous to this, the quantity of stone rubble deposited annually for maintenance purposes on the sea face averaged 85,000 tons for several years. the total length of the breakwater is half a mile. the harbour breakwaters at madras, in a position of extreme exposure, have suffered severe sca damage on many occasions, and large portions have been rebuilt and strengthened from time to time. the original harbour entrance on the east side between the ends of the two curved breakwaters was closed in 1910 on account of the serious ‘‘ range ” in the harbour, due to the position of the entrance and the rapid silting which was taking place. a new en- trance was constructed on the north side of the harbour protected by a sheltering arm.! this arm, as well as the new portion of the breakwater closing the old entrance, is of sloping block work on a rubble foundation protected by 30-ton wave breaker blocks on the sea face. the outer portion of the sheltering arm was destroyed dur- ing a cyclone in nov. 1916, and has since been reconstructed. the sloping or sliced block system of construction is to an increas- inz extent adopted both for breakwater and quay construction, par- ticularly on foundations which are liable to scttlement. harbours.—the bulk of new construction in connection with already existing harbours is in the direction of providing increased depth of water to meet the growing demands of shipping. in many cases of coastal harbours, existing breakwaters, moles or piers have been extended seaward, thus adding to the sheltered area water space of increased depth. in other cases entirely new breakwaters, or other means of shelter, have been constructed and in some already existing channels or sheltered areas have been deepened. important harbour works, particularly in africa, south america and the east, were undertaken in posi- 1 prac, inst, c. e., vol. 190 (1911-2). tions where no artificial protection previously existed, e.g., the african ports of casa blanca, takoradi and killindine and antofagasta in chili. the mersey bar.—the formation and maintenance of a deep- er channel through the mersey bar have been for many years problems of difficulty. dredging by means of powerful sand pump dredgers is carried on almost continuously and the quantity of material removed in recent years from the bar channels and the river approaches averages about 20,000,000 tons annually. since 1907 rubble stone training walls or revetments have been constructed on the east side of the crosby channel and more recently the west side has been similarly treated, but in spite of these training works and intensive dredg- ing the minimum depth of water in the mersey approach chan- nels has not been increased beyond the 26 ft. or 27 ft. at low water ordinary spring tide (l.w.o.s.t.) available in 1907. dock construction in various porrts in great britain, the principal developments in dock ‘con- struction after 1gro took place in the port of london, at liver- pool and at southampton. important works have also been carried out, or were (1926) in progress of construction, at glas- gow, avonmouth, immingham, hull, newport and elsewhere. on the continent of europe perhaps the most important exten- sions have been made at antwerp and rotterdam, and the docks and basins of havre, marseilles, dunkirk, rouen and several italian ports have also been extended. in the east large de- velopments have taken place at calcutta, karachi and bombay, as well as in many of the ports of australasia and japan. port of london.—since the port of london authority was con- stituted in 1909 to take over the docks of the previously existing dock companies and the tidal portion of the thames conservancy jurisdiction, £12,000,000 had been spent up to 1926 on improvements in the port. the ning george v. dock, an extension of the royal albert dock, is the most important addition. commenced in i912, it was opened for traflic in july 1921, the wet dock has an area of about 64 ac., a depth of water of 38 ft., and a total length of quayage of just over two miles. it is connected with the royal albert dock by a passage 100 f{t. wide. an entrance lock, 800 [{t. long by roo ft. wide, with a depth of water over the sills of 45 ft. at trinity high water, gives access from the river thames. a dry dock entered from the wet dock is 750 ft. long and 100 ft. wide, with a depth of 35 ft. over the keel blocks. an interesting feature of the wet dock is the provision of seven reinforced concrete jetties arranged in a line parallel with the south quay at a distance of 32 ft. therefrom. each jetty carries six three-ton cranes. river barges lie between the jetty and the quay wall, and the cranes upon the jetties can deal with goods on the quay or in the ship or barges. the quay walls of this dock are of concrete with vertical faces. the water level of the dock is maintained permanently at a level of 2 ft. 6 in. above trinity high water by means of impounding pumps. this system of raising the dock water level has been adopted since 1910 in the case of most of the london docks with the object of increasing the available depth in the berths alongside the quay walls. at tilbury a new branch dock was constructed as an extension of the old tilbury dock, adding approximately 17 ac. to its water area, a portion of the quay walls of this dock consists of built-up concrete well monoliths, sunk from a level a little below the original ground 194 surface, with mass concrete superstructure, and affords a typical example of modern practice in quay construction in comparatively bad ground.! a river cargo jetty of reinforced concrete sited alongside the deep water channel ae the river at tilbury in 30 ft. of water at l.w.o.s.t. (low water ordinary spring tide) was completed in 1921. this jetty, a novel feature in the port of london, 1,000 ft. in length and 50 {t. wide, is connected with the shore by means of an approach viaduct carrying a railway which extends over the roof of the storage sheds, forming a double-decked superstructure. electric cranes placed on the upper deck unload cargo from ships alongside cither into rail- way wagons or into the transit shed underneath or into barges in the waterway behind. among other improvements carried out by the authority are ex- tensions at the surrey docks and new quays at most of the docks; storage and transit sheds have been modernised and extended, mechanical equipment has been installed for handling such imports as grain and hosen meat, and the river channel has been deepened successively until in 1925 a minimum depth of 27 ft. at low water ordinary spring tide was available from the sea up to the king george v. dock. oil wharves, constructed by private enterprise, have been built in the lower reaches of the river for dealing with the import of oil in bulk. private enterprise is also responsible for im- proved facilities in the port for handling seaborne coal imports. liver pool—the gladstone dock extensions were commenced by the mersey docks and harbour board at liverpool in 1910. these new docks at the north or seaward end of the dock estate constitute the most important addition to the commercial docks of great britain made during the present century. the gladstone dry dock, opened in 1913, was the largest in the kingdom, and at the time of its com- pletion the largest in the world. it is 1,050 ft. long and 120 ft. wide at the entrance, which is closed by a sliding caisson. the depth over the sillis 44 ft.at h.w.o.s.t. (high water ordinary spring tide). orig- inally it was entered direct from the river, but was closed in 1922 during the building of the gladstone wet docks, from whose water area the dry dock would ultimately be entered. the dry dock was constructed so as to be available for use as a berthing dock when not required for dry docking purposes, and its side walls are equipped as quays with cargo cranes and transit sheds. the new wet docks, having a total water area of over 60 ac. are entered from the mersey through a lock 1,070 ft. long and 130 ft. wide, sited so that its cen- tre line is at a small angle with the frontage line of the liverpool docks. at high water ordinary spring tide there is a depth of 48 ft. 9 in. over the sill, and at low water ordinary spring tide 21 ft. 8 inches. the lock is closed by steel gates and is provided with an inter- mediate pair of gates subdividing it into two sections. the new docks contain 2} m. of quayage, to have at maximum high water a depth alongside of 54 feet. treble storey transit sheds with flat roofs are provided on all quays and are equipped with electric cranes both on the quay frontage and on the roofs of the buildings. a new lock go ft. wide connects the gladstone system of docks with the adjoin- ing hornby dock. it was expected that the new docks would be completed about the end of the year 1926. concurrently with these extensions several of the older docks at liverpool have been im- proved or enlarged since 1910, and a new 8o ft. entrance lock has been made at the alfred dock, birkenhead. southampton.—the port of southampton was for many years one of the two principal ports of great britain accommodating the largest class of transatlantic liners, and after the world war the largest of these vessels used southampton in preference to liver- pool. asan indication of the provision which port authorities regard as necessary for traffic of this description, the scheme of dock exten- sions approved by the southern railway co. in 1924 may be briefly mentioned. preliminary work on this scheme was put in progress. all the ship berths at southampton are tidal, and there are no closed wet docks at the port. the extensions were to be constructed on the north side of the river test, to cover a frontage of about two miles. deep water jetties, each 1,000 ft. in length, mainly of reinforced con- crete with a depth of 35 ft. at low water alongside, to be increased by dredging to 45 {t., were to be constructed, graving docks larger than any existing to be built, and a square mile of foreshore re- claimed from the sea. the cost of the work was estimated to be about £13,000,000. since 1910 the trafalgar dry dock at southampton has been en- larged, and is 912 ft. long internally, 100 ft. wide at the entrance, closed by a sliding caisson, and has 35 ft. of water over the sill at high water. the ocean dock, opened in 1912, is an open basin of 16 ac., having a depth of water alongside the quays of 4o {t. at low water ordinary spring tide, the walls being designed for a depth of 45 ft., which can be attained by further dredging. the berths in the ocean dock accommodate the largest transatlantic liners, such as the “ majestic.’”’ the approaches to southampton are dredged to give a minimum depth in the navigable channel of 35 {t. at low water. dublin quay wall.—a quay wall constructed in the port of dublin, 1923-4, is an example of a form of monolith construction which differs from that described above in the case of the tilbury dock walls. reinforced concrete cellular caisson monoliths qo ft. long, 41 ft. high and 28 ft. wide were floated into position and 2 proc. inst. c. h., vol. 215 (1922-3). port engineering sunk in a trench previously dredged to and prepared at the correct level for the foundations. the cells of the monoliths, each of which weighed about 1,200 tons, were afterwards filled in with concrete and gravel and surmounted by a mass concrete superstructure wall (fig. 2). the berth in front of the quay, 35 ft. deep at low water, was subsequently dredged to the required depth. antwerp.—the enlargement and modernising of the port of antwerp on a large scale was undertaken in 1913, but at the com- mencement of the world war a small part only of the new works had been constructed. operations were resumed in 1921. a bassin canal, which is in effect a series of enclosed docks, extending from kruisschans, on the scheldt seven miles below the city to the old dock system higher up the river at antwerp was put in course of construction. the lock at kruisschans is 886 {t. long, 115 ft. wide and has a depth over the sill of 46 ft. at high water. for the con- struction of these works caisson and concrete monolith foundations have been extensively employed. india.—the construction of the king george docks at calcutta, commenced in 1920, when completed in 1927, were to provide ample accommodation for the trade of that port for years to come. one hundred and ninety ac. of water ee eh and 25,000 lineal ft. of deep water quay frontage would be available in the new wet docks. here again well monoliths of large dimensions, sunk through the alluvial material of the ganges delta, were employed for the con- struction of the quay walls and other works. at bombay, a wet dock of 50 ac., commenced in 1905, was opened in 1914. the entrance lock, 750 ft. long and 100 ft. wide, has a depth over the entrance sill of 41 ft. 3 in. at high water ordinary spring tide. the hughes dry dock at bombay, constructed about the same time as the wet dock, is 1,000 ft. long with an entrance 100 ft. in width. at karachi, additional accommodation on a large scale has been made by the construction of deep water tidal wharves, and a similar course has been adopted at rangoon. south africa.—in south africa, the port of durban has been improved by the dredging of the bar, entrance channel and harbour, the normal depth available at low water being now 31 feet. berths alongside the principal open tidal wharves have depths up to 38 ft. 6 in. at low water. at cape town the victoria basin, 67 ac. in extent, provides sheltered berths alongside quays with 36 ft. to 40 ft. water at low water ordinary spring tide, and the breakwater is now (1926) extended. at port elizabeth, breakwaters which will enclose a harbour of 87 ac. were in course of construction in 1926, the available depth in the harbour at low water being 36 feet. both at cape town and durban facilities for the shipment of grain ona large scale have been provided. australia and new zealand.—\n the principal ports of australia where the tidal range is small, open wharves and jetties have been constructed to keep pace with the demands of shipping. at sydney, adelaide and hobart, depths of 35 ft. or more are provided along- side the open quays or jetties. berths having 37 ft. depth at low water are provided at melbourne. in new zealand, auckland and wellington provide for 35 ft. and 36 ft. depth respectively. north america.—on the atlantic and pacific coasts of north america, where the tidal range is usually small and the majority of the larger ports are in situations possessing ample natural protection, e.g., new york and san francisco, the necessary additional accom- modation has been provided by the construction of long and wide jetties or wharves projecting from the shore into deep water along- side which vessels berth. many of these jetties, or ‘‘ docks " as the are termed in new york, have berths upwards of 1,000 {t. in lengt with 35 ft. to 45 ft. depth of water at low water. timber and rein- forced concrete piles are commonly employed for these structures, but sometimes reinforced concrete cylinders are used in place of piles for carrying the superstructure of the jetty. the ballantyne pier at vancouver (fig. 3), completed in 1923, is a good example of recent construction. it is 1,200 ft. in length, 341 ft. wide, with a depth of water in the berths of 45 ft. at low water. the heart of the pier is formed of sand and gravel filling; the outer rows of columns carrying the reinforced concrete superstructure consist of 7 ft. diameter reinforced concrete cylinders, and the inner rows are carried on clusters of piles. two-storey transit sheds, railways and cranes are carried upon the superstructure.? naval doockyards the naval dockyard at rosyth, on the north side of the firth of forth, was in course of construction when the world war broke out. the works had been commenced in 1909 and, although a large part of the accommodation was available for the fleet during the war, construction continued until 1921. the rosyth dockyard occupies a site on the foreshore of the firth above the forth bridge, and comprises a closed basin of 54 ac., open wharves, an entrance lock 850 ft. long, roo ft. wide at sill level, and having a depth of 54 ft. over the sill at high water. there are three dry docks having dimen- sions approximating to those of the jock and depths of 40 ft. over the sills. the lock itself can be used as a dry dock if necessary. in addition to the lock there is a tidal entrance to the basin 115 ft. wide at sill level. concrete well monoliths were used extensively in con- structing the basin and other walls. h. m. govt. decided in 1925 to 2 inst. c. e,. selected papers no. 27 (1925). porter, h semi~portal ee track pai eck “ernie be ue a.c. cylinders tit - 6 wiel ae 2 a 5 aces ° fic. 2.—ballantyne pier vancouver. . reduce the dockyards at rosyth and pembroke to a care and main- tenance basis, and to concentrate all admiralty home dockyard work at the remaining yards at portsmouth, devonport, chatham and sheerness. to meet the growing needs of the nav y, new locks and a dry dock were constructed at portsmouth and completed shortly before the war, the dimensions of which are practically identical with those at rosy th. the largest of the devonport locks and dry docks are of somewhat smaller dimensions. in 1922 the admiralty initiated a scheme for the construction of a great naval dockyard at singapore. the preparatory work which was already in hand was cancelled by the government early in 1924, but with the advent of anew ministry work was resumed at the end of that year. the navy departments of other nations, particularly the united states and japan, have carried out the construction of new docks to meet the growing requirements of their flects. the german naval dockyards at wilhelmshaven, kiel and brunsbiittel, the con- struction or improvement of which had been actively carried on in the years immediately preceding the war, were partially dismantled conformably with the peace conditions, and the naval harbour works at heligoland were demolished under the supervision of the allies (1920-2}. dry and floating docxks dry docks.—in addition to the dry docks which are mentioned in the foregoing paragraphs, large dry docks were completed in the years immediately preceding or following the war in many parts of the world. at esquimalt, victoria, b.c., a dock com- pleted in 1925 has an entrance 125 ft. in width with a depth of water of 4o ft. over the sill, this dock is 1,150 ft. in length. the congella dry dock at durban, opened by the prince of wales in 1925, is 110 ft. in width at entrance and 1,150 ft. long; the depth over the sill is 35 ft. at low water ordinary spring tide. at havre a dry dock commenced in 19t1r was nearing completion in 1926. a novel method was adopted for its con- struction on a foundation of sand and silt. the entire dock structure is built in and upon a huge steel caisson framework 1,132 ft. by 197 ft., which was floated into position over the site of the dock previously dredged to the required depth and there sunk in place. the dock is 1,023 ft. long internally, 125 ft. wide at the entrance, with a depth over the sill of 52 ft. 6 in. at high water ordinary spring tide. the balboa dry dock at the pacific end of the panama canal, opened in 1916, is 1,110 ft. long, 110 ft. wide, with 41 ft.5 in. over the blocks at mean high water. other large dry docks are the boston u.s. navy dock, 1,470 ft. by 114 ft., and that at st. john, new brunswick, 1,165 {t. by 125 feet. floating docks.—the largest floating dock in the world, that at southampton, was opened by the prince of wales on june 27 1924. designed to accommodate the largest ship afloat, it has an overall length of 960 ft., a clear width of 130 ft. 8 in. and a total lifting power of 60,000 tons. the dock can be sub- merged to take in ships drawing 38 ft., and the berth where it is moored is dredged to a depth of 61 ft. at low water ordi- nary spring tide. this dock is of the self-docking type, and is constructed in seven separate sections. -—portland rail tracks or cas: roadway 2 mea renee wr ee 7 a on ae po re “sa nd. ‘e ‘sravel’ fering : =i 195 semi-fortal crane lrack een od en = ) hwost. h ave. ord. l.w. te. ff -———<— === . .- * ' «- | 4, 6 t 29 t a - la een "sr @ bam oon oe a as & am id a ® a bu 4 a e t outer berths. dock and lock entrances.—for closing the entrances of large dry docks steel caissons are used in most modern instances either of the sliding type, as at rosyth and durban, or the floating pattern, of which examples are found at the hughes dock, bombay, havre and the calcutta dock. sliding caissons, although more costly than floating caissons, have the advantage of being more easily and rapidly moved, and, moreover, they are more convenient for the provision of railway tracks and roadw ays over them. steel gates are generally preferred for closing the entrances to large locks and wet docks, although sliding or ship caissons have been adopted in the case of admiralty wet dock entrances, including rosyth and ports- mouth, and at the king george docks, calcutta. the majority of modern steel gates are of the semi-buoyant type provided with buoyancy chambers, which relieve the anchorage at ihe head of the heel post of a great part of the horizontal stress due to the weight of the gate. in such cases the use of rollers and roller paths under the mitre posts is commonly dispensed with, a feature of the design of all modern dry docks, locks and wet docks is the construction of entrances rectangular in form, thus approximating to the box shape of the hulls of modern steamships. for operating sliding caissons both electric and hydraulic power have been employed. for gate opening, machinery of the largest type hydraulic power is preferred in great britain, but electric machinery is employed generally on the continent of europe and in america, as, for instance, for the gates of the panama canal locks. bibliograpily —proceedings of institution of civil engtneers (passim); proc. international navigation congresses (philadelphia, iqi2; london, 1923); brysson cunningham, fzearbour engineering (191 8): mwe dds hunter, dock and lock machinery brie brysson cunningham, dock engineering (1922). cn. gigi porter, horace (1837-1921), american diplomatist and soldier (see 22.116), died in new york city may 29 rg2t. porter, william sydney: see henry, o. portland, ore., u.s.a. (see 22.120), the financial, industrial and distributing centre of the vast columbia river basin, in- creased 24-6 ° in the decade after rgro, to a population of 258,- 288 in 1920, of whom 3,771 were chinese, japanese and indians, 1,556 negroes and 49,778 foreign-born (white and coloured). the census bureau estimate for 1925 was 282,383; the local estimate for the metropolitan area 460,000. annexations of territory increased the area to 63:2 square miles. by the expenditure of $16,000,000 by the federal govt., and $4,300,000 by the city, the entrance to the harbour was im- proved, deepened and protected by two rock jetties, and the river channel was deepened to accommodate vessels of 30-ft. draft at zero stage, from the mouth of the river to portland. by 1925, 54 steamship lines were giving regular service to the prin- cipal ports of the world. the harbour (fresh water, with a frontage of 29 m. within the city limits) had berthing space for roo vessels, 65 ac. of cargo space, 6-5 m. of docks, 50 piers, 4 municipal terminals and many private docks, especially for lumber, grain and flour. during the first five years after the close of the world war, the foreign commerce of the port in- creased by 196% 4 in tonnage, to 1,506,210 tons (1924); and local trade by 114% to 2,872,147 tons. outbound shipments (by rail and water) consisted chietly of lumber, apples, flour, wheat, 196 paper, prunes, barley, canned fruits, salmon (gradually decreas- ing), cascara, copper, hides and wool, doors, ready-made houses (largely to japan), furniture and other articles of woodwork. building permits in the five years following the war amounted to 77,545, representing a value of $116,355,966. the output of the manufacturing establishments within the city limits was valued at $46,861,000 in 1909; $1096,380,000 in 1919; and $154,414,779 in 1923. the commission form of government was adopted in 1913; a permanent zoning commission was estab- lished in 1918; and a housing scheme was adopted in 1919. a system of main streets, park roads and boulevards was laid out and prade crossings were eliminated.