PROCeeoiNGS : PHIIvOSOPHICAL SOCIETY 395 



shelter. The lecturer exhibited a series of tables, curves, and columnar 

 diagrams illustrating the statist cs that have been obtained on these 

 subjects in England, and the decrease in infant and child mortality 

 that has followed improvements in the conditions. An item of particu- 

 lar interest in this series of data dealt with the effect of diminished 

 infant mortality on the death rate of the community as a whole. It 

 has been urged by many that the saving of infant lives works directly 

 against that process of natural selection which is assumed under the 

 conditions of civilized life to continue to weed out unlit individuals 

 early in life. If this were the case it is to be expected that com- 

 munities having a high infant mortality should show a decreased 

 mortality in part or all of the subsequent age periods. The statistics 

 show that exactly the contrary is the case. Individuals in a com- 

 munity with high infant mortality have a smaller expectation of life 

 throughout their entire period of life than do individuals in a com- 

 munity with low infant mortality. In other words, any action of 

 natural selection that continues is so small a factor under present condi- 

 tions that it does not appear on the curves. The same preponderant 

 effect of evil environmental circumstances persists in the successive 

 years of life as was evidenced in infancy. 



The lecturer then showed a series of diagrams illustrating the various 

 agencies that have been developed in England to remedy the conditions 

 described. 



The discussion was participated in by Drs. Wall, J. A. Foote), 

 G. M. KoBER, Livingston Farrand, and T. A. Williams. 



Robert B. Sosman, Corresponding Secretary. 



PHILOSOPHICAL SOCIETY OF WASHINGTON 



The 826th meeting was held at the Cosmos Club, Dec. 20, 1919, with 

 President Sosman presiding and 34 persons present. The minutes of 

 of the 823rd and 824th meetings were read 'n abstract and approved. 



The first paper was by Mr. J. Warren Smith on Predicting minimum 

 temperatures,^ and was illustrated by lantern slides. 



This paper was a mathematical discussion of the relation between 

 the relative humidity in the late afternoon and the variation of the 

 minimum temperature during the coming night from the afternoon 

 dewpoint temperature, when radiation conditions prevail. The study 

 shows that there is a well defined relation which can be expressed by the 

 curve for a parabola. This curve can be constructed by the "star 

 point" method of curve fitting instead of by the more tedious well 

 known least square method. 



The equation used is written v = x -{- by -\- cz \n which v is the 

 variation of the minimum temperature from the evening dewpoint; 

 h is the evening relative humidity, and c is the square of the relative 

 humidity, x, y, and z are the three unknowns, which are evaluated 



^ Full publication in Monthly Weather Review Supplement, No. 16, pp. 6-19. 

 1920. 



