MENDEL'S PRINCIPLES APPLIED TO WHEAT HYBRIDS. 877 



the remainder were identical with the female parent, thus showing that 

 the flowers had been self-fertilised. This is made evident by the results 

 below. 



" In each case one of the parents was of the club type (Triticitm corn- 

 pa c turn), the other of the common type (T. vulgare). As above stated, no 

 important variations occurred in the first generation except as noted 

 below ; but when the heads appeared on the second generation, a remark- 

 able state of affairs was seen to exist. At the first glance it appeared that 

 each of the hybrids had split up into all sorts of types. But closer 

 inspection showed that in every case but one, which is noticed later, the 

 forms in each plat were simply combinations of the characters of the 

 parent forms. Further inspection revealed the fact that in plats of 

 similar breeding exactly the same types were present. This suggested 

 the idea that perhaps a hybrid tended to produce certain definite types, 

 and possibly in definite proportions. Accordingly, all the hybrid plats 

 were assorted into types, and the proportion of each type determined. 

 "When these results were classified, they confirmed the above suggestion ; 

 and if similar results are shown to follow the crossing of other groups 

 of Wheats, it seems possible to predict, in the main, what types will result 

 from crossing any two established varieties, and approximately the pro- 

 portion of each type that will appear in the second generation. With the 

 exceptions already referred to, the second generation consisted of the two 

 parent types and of all the intermediate types possible between them. For 

 instance, when one parent had long bearded heads, and the other short 

 beardless heads, the plat could be divided into six types : two of these had 

 long heads like one of the parents, two others short heads like the other 

 parent, and two were intermediate ; and one each of these three groups 

 had beards, while the others had none. In some of the crosses the parent of 

 the common form had velvet chaff, but no beards. Here a similar set of six 

 types appeared, the velvet replacing the beards. When one parent had velvet 

 chaff of dark brown colour, twelve types were possible, and were actually 

 found ; six of these were similar to the six above, and the six others were 

 like them except that they had brown chaff. As might be expected, many 

 plants occurred more or less intermediate between these types, and it required 

 considerable care to assort the types satisfactorily. 



" Appended hereto are tables showing the percentages for all the plats 

 that were assorted with sufficient care to be included here. To show the 

 close agreement frequently found in the proportions in which the same 

 types occurred in different plats, let us examine the figures for three plats 

 of the same breeding. Type I., which was like the female parent, con- 

 stituted 6 per cent, of one plat, 66 per cent, of another, and 6*9 per cent, 

 of another. The percentages for the other types were as follows : II., 

 17-7, 17-9, and 19'1 ; III., 12-7, 12-4, and 13-1 ; IV., 40"5, 38-8, and 37'2 ; 

 V., 5*7, 4*9, and 7*3 ; VI., which was like the other parent, 16*4, 19*5, and 

 19*9. The agreement is not usually so close as this, but it is fair to 

 assume that the discrepancies are partly due to the small number of 

 plants in each plat. They are also partly due to the fact that not a few 

 plants were more or less intermediate between the types, and these were 

 sometimes placed in one group and sometimes in another. 



