i68 NATURAL SCIENCE, March. 1897. 



an instinctive sense of direction cannot always be the agency- 

 employed ; and, as in no one case is the exact route of the animal's 

 return known, the assertion so often made, that a special sense of 

 direction is required to explain the facts, cannot be justified. 



Two very interesting chapters, near the end of the volume, are 

 those entitled " Are Acquired Characters Inherited," and " Modifica- 

 tion and Variation." In the first of these the bearing of the whole of 

 the phenomena of instinct on the vexed question of inheritance is 

 pointed out, and the conclusion is reached, that although there is Httle 

 or no satisfactory evidence of the transmission of acquired modifica- 

 tions — that is of habits, or their effects on the organism, as opposed 

 to instincts, yet there are many curious facts which seem to indicate 

 some connection between congenital and acquired characters. What 

 this connection is, the chapter on Modification and Variation attempts 

 to show. 



Modification of the individual by the environment, whether in the 

 direction of structure or of habits, is universal and of considerable 

 amount, and it is almost always, under the conditions, a beneficial 

 modification. But every kind of beneficial modification is also being 

 constantly effected through variation and natural selection, so that the 

 beautifully perfect adaptations we see in nature are the result of a 

 double process, being partly congenital, partly acquired. Acquired 

 modification thus helps on congenital change by giving time for the 

 necessary variations in many directions to be selected, and we have 

 here another answer to the supposed difficulty as to the necessity of 

 many coincident variations in order to bring about any effective 

 advance of the organism. In one year favourable variations of one 

 kind are selected and individual modifications in other directions 

 enable them to be utilised ; in Professor Lloyd Morgan's words — 

 " Modification ^5 such is not inherited, but is the condition under which 

 congenital variations are favoured and given time to get a hold on the 

 organism, and are thus enabled by degrees to reach the fully adaptive 

 level." The same result will be produced by Professor Weismann's 

 recent suggestion of " germinal selection," so that it now appears as 

 if all the theoretical objections to the " adequacy of natural selection " 

 have been theoretically answered. 



Biologists owe a debt of gratitude to Professor Lloyd Morgan for 

 this most interesting and suggestive volume. It exhibits all the 

 clearness and philosophical acumen which characterise the writings of 

 the author, and although in his desire to be impartial he has sometimes 

 suggested difficulties which are more apparent than real, yet the work 

 is on the whole an admirable introduction to the study of a most 

 important and fascinating branch of biology, now for the first time 

 based upon a substantial foundation of carefully observed facts and 

 logical induction from them. 



Alfred R. Wallace. 



