40 POTATO DEVELOPMENT WORK IN WISCONSIN. 



the composite race from which it was derived. This improvement 

 may be in the form of a more productive plant, or one whose 

 tubers are more uniform in shape and size, or the eyes may be 

 fewer and shallower, or any other characters which seems more 

 desirable. In other words some of the fluctuating variations oc- 

 curring within a variety may be heritable factors which have 

 heretofore been assumed to be now heritable. A word of caution 

 should be given 10 the selection with respect to the isolation of 

 types which are apparent departures from the true type of the 

 variety. In my opinion there can be but one excuse for the isola- 

 tion and that is that it is decidedly superior to the parental type. 

 It seems to me that in the long run much more can be accom- 

 plished by selecting strains which most uniformly conform to the 

 varietal type. It is only by rigidly adhering to this principle 

 that we can hope to make any progress in the standardization of 

 our commercial varieties. To intelligently improve a variety it 

 is necessary that one should have an intimate knowledge of its 

 normal type. 



SELECTION METHODS. 



The improvement of the potato by selection may be accom- 

 plished in either of two ways viz. The "hill selection " and the 

 "tuber unit" methods. They only differ in one respect and that 

 is in the way in which the first selection is made. 



HILL SELECTION METHOD. 



The usual practise in making hill selections is to carefully ex- 

 amine the plants in the field when they are fully grown and mark 

 a'.l such as most nearly .represent the ideal type of the variety and 

 that are vigorous and haalthy. Prior to harvesting, these hills 

 should be dug by hand and the tubers carefully inspected. Re- 

 ject the progeny of all plants which do not show a large propor- 

 tion of marketable tubers closely adhering to the varietal type 

 and of fairly uniform size. A number should be given to each 

 selection and a record made of the number of marketable and un- 

 marketable tubers after which the tubers from each selection 

 should be stored in separate sacks or other receptacles. The fol- 

 lowing season plant your selections in such a way that each will 

 occupy a separate row or portion of a row, and going still fur- 

 ther, plant each tuber of the selection so that its plants may be 

 readily identified from those of every other tuber of that number. 



