EXPERIMENT STATION REPORT 263 



almost perfect gradation of type and color to either parent. 

 (Plate XXIII). 



"Warren-Boston Marrow" (7/2 I). ("Warren" type, with- 

 out a navel used). The fruits from this selected type are vari- 

 able but the absence of navel of the "Warren" is constant. 



"Warren-Boston Marrozv" (7/2 I). ("Warren" type with the 

 navel used). Some of the fruits in the progeny were with, and 

 others without, the navel, but wherever the navel was present 

 it was small (Plate XXIII). 



"Delicious-Golden Hubbard" (3/4). ("Golden Hubbard" 

 type, long, used). Not one of the fruits in the progeny was 

 long. It is interesting to note that, while the mother fruit .was 

 orange, the entire lot of the progeny was like the male parent in 

 color and shape. 



"Golden Hubbard-Hubbard" (4/1). ("Hubbard" type, large, 

 used). In this case, the size of fruit bred true but the dark green 

 color of the mother parent did not appear; all were orange like 

 the male parent. 



These few instances of selection of winter squashes are suffi- 

 cient to convince us that there is little hope for immediate 

 results. We are now ready to see the behavior of a few crosses 

 in the second generation or the third year after the cross, 



"Hubbard-Boston Marrow" (1/2 II). The general shape of 

 the fruits is nearly uniform. Of the fourteen fruits ten are 

 striped, three solid green and one is orange, showing that the 

 majority are beginning to acquire uniformity. 



"Hubbard-Boston Marrozv-Dclicious-G olden Hubbard" (1/2// 

 3/4 II). /Most of the fruits are long. Of the seventeen fruits 

 in the progeny ten are green, six green with orange spots and 

 one solid orange. This cross also shows that type is tending to 

 acquire uniformity. 



"Delicious-Hubbard" (3/1 II). The crop is uniform in both 

 shape and color. 



"Delicious-Golden Hubbard" (3/4 II). All the fruits resem- 

 ble the "Golden Hubbard" in shape, and nearly all in color. 



The four cases quoted above indicate that, by continual selec- 

 tion, it will be possible to establish uniformity among crosses 

 of winter squashes as well as among summer squashes, demon- 

 strated the last two years. 



It is here proper to speak of the variations that occur from 

 time to time among commercial varieties. The "Cocoanut," 

 which is a well-established, smooth-sized fruit, produced this 

 season one plant, among forty or fifty, bearing fruits three to 

 four times the ordinary size. In 1907, one fruit bred within the 



