88 REPORTS OF INVESTIGATIONS AND PROJECTS. 



fied, increased, or diminished, and there is evidence that such modifications 

 are sometimes inherited. Thus Castle has shown that the extent of the 

 pigmented area in rats may be varied in an inheritable fashion by selection of 

 slight variations and, beginning with a scarcely recognizable trace of syndac- 

 tylism, I have succeeded in getting very exaggerated forms of this condition. 

 On the other hand attempts, in other cases, to increase or diminish character- 

 istics (i. e., certain color-characters) by selection have not yet met with suc- 

 cess. This whole subject of the modifiability (and particularly the inheritable 

 modifiability) of characters deserves thorough investigation. Here lies the 

 crux of the controversy between the Darwinian "selectionists" and the De 

 Vriesian mutationists. 



Of great importance in this regard is the question whether the soma can 

 modify the germ-plasm in a detailed way. Guthrie, two years ago, adduced 

 evidence for this view, but his results have not been confirmed either subse- 

 quently in his own publications nor by Dr. Castle working with rats. We 

 are making experiments with poultry that should test this doubtful matter 

 thoroughly. 



Meanwhile we are extending our knowledge of the extent of modification 

 that characteristics undergo in nature, Drs. Harris and Shull and R. H. 

 Johnson and W. L. Tower having obtained extensive series of data from 

 various species of plants and insects. These studies, as well as those on 

 selection, lend support, in many cases, to the view that natural characteristics 

 in a state of nature undergo a progressive change in a definite direction. 

 Evolution is proceeding in consequence of internal changes in the germ-plasm 

 that are doubtless controlled by external conditions. 



Extensive studies on the effect of external, particularly nutritive, conditions 

 on the development of the form and structure of plants were begun by Dr. 

 E. F. Transeau at this Station, but they have not yet been reported on. The 

 results of starvation or semistarvation were very marked. 



THE RELATIONS OF CHARACTERISTICS. 



In any organism characters do not exist alone but are related to other char- 

 acters and to the external world. This fact is the basis of the phenomena 

 of correlation of characters and of the elimination of unfit characters by the 

 selective annihilation of individuals carrying them. We have seen how the 

 whites and the solid black of poultry make them conspicuous and especially 

 liable to be killed by crows. Dr. Harris has published accounts of a series of 

 observations directed toward determining whether the eliminated differ in a 

 given particular from the surviving. The results are usually negative, for 

 the characters considered. 



This is, then, the situation in which this Station finds itself at the end of 

 the first half -decade. The attack on the problem of the organic characteristic 



