jpoo 1 .] IMPROVEMENT OF THE POTATO. 405 



He concludes that the vigor of the plant may be maintained and 

 even increased by selection ; but that the experiment has not tended 

 to increase the yield of the varieties used, for the reason, he says, 

 of the continued cutting of the best tubers. 



Bolley (12) has for a number of years carried on an investiga- 

 tion concerning the use of large and small tubers from the same 

 hill. Bud end pieces of equal weight, cut to one eye piece were 

 used, and all details of the experiment very accurately controlled. 

 From three years' observations, he concludes that: "In planting 

 equal weight pieces from small and large tubers of the same vine, 

 there will not be a sufficient difference in favor of the one or the 

 other size of potatoes to be noticeable under farm methods, pro- 

 vided all are normally mature." This shows that selection of large 

 potatoes from the bin will not necesssarily do anything toward im- 

 proving the crop. He concludes also from his general observations 

 on his stock, that variations in forms, size, roughness of skin, mal- 

 formations, number of tubers, etc., are produced to some extent 

 though influenced greatly by seasonal conditions. From this he 

 concludes that selecting from high yielding plants should improve 

 the yield. Later ( 13), he reports his first conclusions to be justified 

 by more data. 



Fruwirth (43) criticises Bolley 's work stating that he had 

 found that large tubers were more likely to produce large tubers 

 than were small tubers even from the same vine. He gives but 

 few data warranting this conclusion, and I think that he overlooks 

 Bolley's qualifying statement that "all must be normally mature." 



In 1899, Remy (85) saw a part of the error* in the work of 

 former German and French chemists who used physical methods 

 for determination of specific gravity, and estimated their starch 

 from this by Maercker's tables. He found that a great number of 

 the tubers had hollows in their centers which seriously vitiated his 

 results. With the idea of correcting this fault, he selected good 

 smooth potatoes weighing about 2oog. each and used about 7<3g. 

 from the bud end for planting, while he determined the starch 

 gravimetrically in the remainder by the use of Fehling's solution. 

 The results for 1899 an d JQOO show no definite relations between 

 the nature of the seed tubers and the progeny either in dry matter 

 or in starch content. This is the most exact experiment up to this 

 time, although even the chemical method for starch determination 

 leaves much to be desired. The starch fluctuations were rather 

 narrow, however, running in 1898 seed from 76.0 percent to 81.6 

 percent calculated to the dry substance; and two crops are hardly 

 sufficient to settle this question. 



From 1899 to 1901 appeared the notable contributions of 

 Fischer (33, 34, 35) to this subject. He showed that there was 



*See writer's table of fluctuations in nitrogen content for further error. 



