NOVEMBEE 15, 1907] 



SCIENCE 



657 



light. A few months ago, I think it would 

 have been scarcely an exaggeration to have 

 said there was none. It is, however, im- 

 possible not to recognize that the striking 

 experiments lately published by Tower may 

 be a positive contribution to this part of 

 the inquiry. We can scarcely imagine that 

 changes in temperature or in moisture are 

 the great or chief efficient causes of natural 

 variation; still the fact that in Tower's 

 experiments such artificial changes in con- 

 ditions appear to have effected a modifica- 

 tion in the germ cells of the potato beetle 

 {LepUnotarsa decem-lineata) and to have 

 permanently deflected the offspring into a 

 recessive line, must be- allowed weight 

 in future discussions of these phenomena. 

 Many points in that fine piece of work 

 still remain to be cleared up, but a very 

 remarkable beginning has been thus made. 

 It is, perhaps, scarcely necessary to add a 

 warning that though the response to 

 change of conditions may have been direct, 

 it must not be hastily concluded that the 

 response is adaptive. The appeal to direct 

 responses so common in evolutionary dis- 

 cussions of thirty years ago, was made to 

 account for the complex adaptations of 

 organism to environment. It is the total 

 want of any evidence supporting that ap- 

 peal which has driven most of us to dis- 

 believe in the reality of any such claims, 

 and there is nothing in the new evidence, 

 I think, which should shake the attitude of 

 resolute agnosticism which we have thus 

 been led to adopt. 



Similar reflections apply to another very 

 curious instance of genetic change in- 

 duced by more violent means. MacDougal 

 states that by injecting zinc sulphate into 

 the ovary of Raimannia he caused the 

 plant to produce seeds which became small 

 and depauperated plants, destitute of the 

 ciliation characteristic of the parent spe- 



cies. These, in their turn, transmitted the 

 new character to their descendants. 



The facts which I have referred to aa 

 helping to limit our view have been drawn 

 from the behavior of a considerable range 

 of characters and, as I have said, there are 

 strange elements of similarity common to 

 all. Respecting two very important 

 classes of characters we still remain in 

 almost total ignorance. Some years ago in 

 attempting a provisional survey of varia- 

 tions I distinguished a special group of 

 phenomena as meristic, that is to say, be- 

 longing to those occurrences by which 

 division and repetition are effected in ani- 

 mals and plants. Obvious as the meristic 

 differences are, we know very little as to 

 the system followed in their inheritance. 

 Only one case is clear, I believe. Farabee 

 has shown that the peculiar condition of 

 the human digits in which the fingers and 

 toes have only two phalanges each, behaves 

 as a simple dominant. Dr. Drinkwater 

 has very kindly sent me lately a table 

 which he will shortly publish, showing 

 exactly the same thing in an English 

 family. In his family, as in Farabee 's, the 

 affected members were of very short 

 stature. I can not at all readily conceive 

 how any ferment or other transmissible 

 substance can be supposed to be responsible 

 for such a variation as this. It is true 

 that the attacks of gall-flies or of fungi 

 may excite branching, or proliferating cell 

 division in plants, and we may have to 

 suppose that a poison can have this effect. 

 Perhaps we may also imagine that the fine 

 division of the hair follicles in Angora rab- 

 bits or Merino sheep may be due to the 

 want of some substance which in the nor- 

 mal type inhibits or checks this excessive 

 siabdivision, but if we are to bring the two- 

 phalanged digits into line with the rest of 

 these observations we shall have to make 



