BIRDS AND EVOLUTION 381 



liver. The same was true in regard to the kidneys, pointing 

 to a greater toxicity of fish and insect as compared with flesh 

 and vegetable food. 



A flesh-diet proved the best as regards growth, and the 

 best even in adult life, though less markedly so. Insects were 

 nearly as nutritive as beef during growth, but in adult life 

 they proved inferior. Fish and vegetables were less well 

 suited for the younger stages, but in adult life vegetables 

 proved only slightly inferior to beef. Laying began earliest 

 and was most abundant in the fish-fed birds. 



Magnan points out that the modifications brought about 

 in the ducks by each kind of diet correspond in general to 

 the features seen in birds whose natural diet is like that in the 

 experiment. This would indicate that diet may have a 

 considerable influence on the character of the food-canal in 

 the individual bird, but behind this there is probably the 

 fact that in the course of selection that kind of food-canal has 

 been established which is best adapted to the customary 

 diet. 



A very important contribution to the question of the 

 transmissibility or non-transmissibility of individually ac- 

 quired somatic modifications has been made by Professor 

 M. F. Guyer and E. A. Smith (1918). It is quaintly subtle, 

 but it is more value than much discussion. Its introduction 

 here is justified by its importance and also because the 

 corpus vile was the fowl. Lens tissue from rabbits and mice 

 was made into an extract and injected into fowls. It excited 

 in the blood of the fowl the production of specific anti- 

 bodies, that is to say, chemical counteractives to the 

 deleterious intrusions. Now, if the serum of the fowl's 

 blood be injected into pregnant rabbits, the anti-bodies may 

 attack the lenses of the unborn young, producing a lique- 

 faction, apparently dissolving a fibrous protein in the lens. 

 The reaction is not invariable, however, since the majority 

 or even all of the individuals of a litter may escape the cell- 

 dissolving (cytolytic) anti-bodies ; and even when an in- 

 dividual embryo is attacked one eye may escape. This is 

 difficult to explain. No effect was observed in the mother 



