390 



NATURE 



[March 23, 192: 



importance, he arrives at the conclusion that there is 

 a direct relation between it and a standardised death- 

 rate. This is confirmed by some calculations made 

 by Mr. Finch. Applying the death-rates at each 

 quinquennial period of age to a standard population 

 adjusted so as to be in arithmetical progression, for 

 which purpose the mean population for the decade 

 1 89 1 -1 900 was adopted, the errors appeared to be 

 remarkably small. 



Dr. Farr held that density of population and death- 

 rate were closely connected. The difficulty has arisen 

 in applying the two to London, which possesses a 

 greater absolute healthiness than its density would 

 suggest. May we not infer that this is partly due 

 to good management and sanitary conditions ? 



It is less easy to get trustworthy generalisations 

 where the numbers are small. Dr. Brownlee suppUes 

 a valuable series of tables of the numbers living and 

 the expectation of life in selected healthy and unhealthy 

 districts for use in calculating the death-rate from 

 various diseases. He infers from them that persons 

 who died at the age of fifty-one years in the average 

 environment might have had a life of seven years 

 longer in the mean had they lived in the country. 



Proceeding to the consideration of the effect of 

 particular diseases. Dr. Brownlee takes (i) Phthisis. 

 The age at which phthisis causes death is shown to 

 vary greatly in different districts. (2) Sarcoma and 

 cancer. Here for an equal number of deaths the age 

 at death is shown to be later where the conditions are 

 healthier. (3) Valvular disease of the heart. This 

 seems to behave in much the same way. (4) Diabetes. 

 Here, whether a person lives in a rural district or in a 

 county borough, the commonest age at death is the 

 same. (5) Nephritis. This is much less prevalent in 

 rural than in city districts. (6) Pneumonia. Deaths 

 from this are least in rural districts. Care has to be 

 taken in deahng with the three elements of the 

 problem, age, environment, and disease. 



In pt. 2, relating to the mathematical treatment 

 of the subject. Dr. Brownlee seeks to give directions 

 for calculating life-table data by short and easy 

 methods. Those desirous of making inquiry into 

 health conditions will, however, have to bear in mind 

 the many pitfalls that they may meet in so delicate 

 an investigation. Dr. Brownlee's authoritative and 

 suggestive report will enable them to avoid the 

 danger of hasty conclusions. 



T' 



'HE possibilities of the quantitative representation 

 of geographical data as regards population dis- 

 tribution are discussed in some detail by the originator 

 of a new method, Mr. S. de Geer, in the Geographical 

 Review for January. Mr. de Geer has already ap- 

 plied the method in the recently published atlas of 

 the distribution of population in Sweden. In the 

 ordinary map the position and size of cities and 

 smaller centres are shown by dots. A further develop- 

 ment is to show relative density by shading or colour- 

 tints of varying depths. The chief defect of such 

 maps is that, as a rule, they show only the average 

 over large areas such as counties or parishes. 



The dot method of Mr. de Geer offers the possibility 

 of combining a clear representation of situation and a 

 mass of population within wide hmits. The dot 

 represents a unit of population of fixed value ; the 

 larger the scale of the map the smaller the unit. On 

 a scale of i : 100,000 a dot might represent ten 

 persons ; on one of i : 80,000,000 perhaps a million 

 persons. The unit-dots are considered as small 

 spheres, and should be shaded as such, but this 

 involves expense and difficulties in printing ; they are 

 therefore drawn soHd black. Small towns are shown 

 by groups of dots arranged in squares, rectangles, or 

 other figures corresponding roughly with the extent 

 of the settlement. Such regular arrangements at 

 once differentiate urban centres from rural com- 

 munities. Large centres cannot well be shown by 

 dot-nets because of the space required. Urban 



Population Maps. 



populations about a certain number, varying with 

 the dot-unit chosen, are shown by large spheres the 

 volume of which is proportional to the unit-dot 

 and decided by the population of the centre. Thus 

 in the Swedish map the unit-dot, representing 100 

 persons, has a radius of 0-57 mm., and the sphere 

 representing Stockholm (371,000 inhabitants) has a 

 radius of 8-9 mm. These large spheres are shaded 

 to give a spherical appearance. As the quantitative 

 value of the larger sphere is not readily estimated, it 

 is expressed in units printed on or beside it. 



Much geographical judgment must be used in 

 the placing of the dots, especially in rural districts. 

 The population of isolated farms and small hamlets 

 has to be gathered into groups of 100 if that is the 

 unit chosen. The dot is placed either at that place 

 with more than halt this number of inhabitants or, 

 if there is no such place, near the centre of gravity 

 of the group. Due regard must also be had to the 

 density of neighbouring groups, particularly near 

 administrative boundaries. The map is further 

 improved by tints of colour distributed to show 

 relative density of population. Mr. de Geer rightly 

 claims that such a population map has many practical 

 applications in questions of the readjustment of 

 administrative boundaries, of the establishment of 

 public institutions, of lines of communications, of the 

 location of educational facilities, of the stationing of 

 officials, and in other directions. 



The International Fishery Investigations. 



T' 



'HE International Council for the Exploration of 

 the Sea met at Copenhagen in July last, and 

 the official account of the proceedings is now avail- 

 able. An unofficial report, with some criticisms, has 

 also been pubUshed by M. Ed. le Danois. At this 

 meeting Belgium, Denmark, Finland, France, Great 

 Britain, Holland, Norway, and Sweden were repre- 

 sented, and negotiations are in progress for the 

 inclusion of Spain, Portugal, Esthonia, and Lettonia. 

 The Governments of Canada, Newfoundland, and the 



1 Rapports et Proces-Verbaux des Reunions : Conseil International, 

 Exploration de la Mer, vol. 27, Copenhague, December 1921. 



Notes et Memoires, No. 11, Office scientifique et technique des Pfeches 

 Maritimes, Paris, December 1921. 



NO. 2734, VOL. 109] 



United States have meanwhile adopted a joint scheme 

 of oceanographical investigations, and contemplate 

 '' establishing contact " with the European organisa- 

 tion. 



The official report summarises the proceedings of 

 the council, the sections, and committees ; the latter 

 relate to investigations on the herring, cod and had- 

 dock, plankton, hydrography, limnology, statistics, 

 the Atlantic slope, the Baltic, and the plaice. Pro- 

 grammes of the researches contemplated in each of 

 these subjects are given, and there are indications of 

 the share taken by each country and of the Umited 

 progress that has been made. So far little has been 

 pubhshed. M. le Danois's unofficial report is, in part. 



