SECT. 6] OF THE EMBRYO 877 



were in general agreement with the later independent results of 

 Murray. Dunn found that eggs were two or three times as variable 

 in their evaporation-rate as in their original fresh weight. If, then, 

 the rate at which an egg loses weight is regarded as an index of the 

 permeability or porosity of its envelopes, then the shells of eggs 

 must be more variable than the sizes of the eggs themselves. This 

 was exactly what had been found by Curtis for the weights of the 

 shell, the coefficients of variation of shell- weight/egg- weight being 

 I0-43/6-36. In sum, loss of weight, said Dunn, was to be regarded 

 as an individual character, like weight, length, breadth, shell-weight, 

 yolk-weight, etc. In his second paper, he studied the relation of egg- 

 size to weight loss, and concluded that larger eggs, though they lose 

 more actual weight, lose an appreciably smaller proportion of their 

 weight than smaller eggs. The larger eggs can apparently better 

 conserve their moisture-content, but this is due wholly to their rela- 

 tive surface. The shells of the larger egg, however, were somewhat 

 less porous, for they lost less weight per unit of surface area. This 

 shell-difference was the subject of the third paper. Analyses made 

 on the shells of eggs which had shown high and low rates of evapora- 

 tion respectively revealed no difference except in the shell-weight, 

 thus: 



Egg- Shell- Shell- Calcium Magnesium 



Rate weight weight weight oxide oxide 



of loss (gm.) (gm.) (%) (%) (%) 



High 65-22 5-399 8-28 52-01 1-52 



Low 55-38 5-258 9-49 52-44 1-52 



But, on the other hand, the rate of evaporation was correlated very 

 closely with the number of pores (see Rizzo, p. 722) in the shell, 

 and this was indeed quantitatively much the most important factor. 

 Later Dunn investigated the relations between evaporation-rate and 

 hatchability. He found that under constant conditions the rate at 

 which a normal fertile egg loses weight in the first 7 days of incubation 

 played little or no part in determining the subsequent fate of the 

 embryo contained in it. Nor was the weight loss during the 2nd 

 week correlated in any way with the hatchability. On the other 

 hand Romanov found that the optimum hatch occurs at 60 per cent, 

 humidity. He weighed a large number of embryos from eggs kept at 

 40 and 80 per cent, humidity, and concluded that the latter condition 

 gave slightly heavier embryos than normal although their percentage 

 dry weight was less than usual. At high humidity the percentage ash 



