tl. L D. GRAHAlVt 21 i 



Cross IX. 



The next cross Gunji resulted in the production of three types of plants 

 distinguished as Red grained, Yellow grained, and White grained, respectively. 

 Taking the families in which all three colours occur we find the total Red, 340, 

 Yellow, 96, and White, 139. This approximates to a 9 : 3 : 4 ratio fairly closely, 

 the expectation on that supposition being Red, 325 ; Yellow, 105 ; White, 140. 

 The red grained plants proved in further cultures to be of three types, the 

 yellow of two and the white of one type. On crossing a number of the whites 

 with one of the pure yellows, however, the whites proved to be of at least two 

 kinds, namely, those which gave a red grained Fi plant and those which gave 

 a yellow grained Fj plant.* This leads to the assumption that some of the 

 whites are really undeveloped reds and only require the addition of yellow 

 to cause the red colour to develop. The complementary factors that compose 

 the red colour are separate, one being contained in the white grained planf and 

 the other in the yellow grained. When both of these combine the red colour is 

 produced. The simplest solution suggested by Mr. H. Martin Leake to whom 

 I desire to tender my thanks for the suggestion is one involving only two 

 factors R (red) and Y (yellow). The simplest case of such a combination is 

 one in which RrYy will give 12 Red, 3 Yellow, 1 White, supposing red 

 completely masks yellow. If we suppose that Red can only develop in the 

 presence of yellow then RrYy gives 9 Red, 3 Yellow, 4 White. 



The actual numbers are given in the table on page 212. 



* Artificial cross-pollination. Juar presents few difficvilties with regard to artificial cross- 

 pollination. The inflbrescence can be mutilated very considerably without apparently interfer- 

 ing with the fruit development of the remaining flowers. The fruits also develop freely inside 

 protecting envelopes. The method employed was to select a plant in which the panicle was 

 just emerging from the leaf sheath. In the afternoon all the topmost branches, with the excep- 

 tion of three, were removed ; at the same time the lower branches were removed to expose 

 the rachis which forms a useful support on which to fasten the protecting envelope. The 

 removal of the lower branches further prevented the chance of pollen from the lower flowers 

 adhering to the mouth of the envelope during subsequent examination of the protected flowers. 

 All the flowers were removed from the three top-most branches except one near the tip of each 

 branch and one near the lower end of the branch. The glumes were thereafter gently forced 

 apart by holding the awn with forceps and pressing the glumes laterally. The flower was thus 

 easily castrated and was then protected in the usual way by an envelope. When the stigmas 

 were mature — this being ascertained by removing the envelope, when, if the stigmas are 

 mature, the glumes are found to be open with the stigmas extruding— pollen was applied. It 

 is necessary to make the examination early in the morning, though castrated flowers in 

 envelopes remain open for a longer time than exposed flowers. After the grain had set the 

 bottom of the envelope was torn off, the remaining portion serving the double purpose of a 

 distinguishing mark and a protection from birds. Control flowers showed that there was little 

 danger of accidental self-pollination during the process of castration if ordinary precautions 

 were taken. 



