BIOCHEMICAL STUDIES OF HYBRIDS 321 



Since flower extracts yielded somewhat different two- 

 dimensional chromatographic patterns than did the leaf extracts 

 which were used in the earher study, some additional species-specific 

 compounds are now available. Other classes of compounds, detect- 

 able by other methods, are currently being investigated (for example, 

 the anthocyanins of lower stems differ in Baptisia species). Figure 

 15-3 illustrates in comparative fashion the chromatographic patterns 

 upon which an analysis of the "tri-hybrid" population was based. 

 Chromatograms of individual representative hybrid types are also 

 illustrated. 



When such a large pool of useful compounds is available, it is 

 possible to extend an analysis of hybridization to include the degree 

 and direction of backcrossing. As a working hypothesis one may 

 assume that a hybrid A x B contains approximately the sum of the 

 constituents of A and B, and a backcross to B should have the com- 

 pounds of B and approximately half of those of A. In a real situation 

 the exact composition of A x B cannot be predicted because of lack 

 of knowledge of the mode of inheritance of the specific compounds. 

 Neither can the chemical make-up of backcross types be predicted, 

 and in this latter situation, segregation of genetic factors in the hy- 

 brid parent provides for more individual variation among backcross 

 types than in the hybrids themselves. However, such information can 

 be acquired empirically by analysis of the hybrids of a large popula- 

 tion or many populations. 



To illustrate, in Fig. 15-4 data from the plants of the tri- 

 hybrid group, ten plants of each species from pure populations plus 

 seventeen additional B. leucantha x B. sphaerocarpa hybrids are 

 plotted. Points in this graph for hybrids or derivations include both 

 species-specific compounds and those compounds shared in common by 

 the two species concerned. This type of plot makes it difficult to draw 

 conclusions from the B. leucophaea X B. sphaerocarpa hybrids 

 because these two species share a rather large number of compounds. 

 In contrast, B. leucantha x B. sphaerocarpa hybrids contain mostly 

 species-specific compounds. The hybrid types in both situations cluster 

 at an angle near 45°, indicating, in the case of B. leucantha x B. 

 sphaerocarpa hybrids that most of these plants are truly Fi hybrids. 

 Two plants, 47 and 11 (note arrows), fall suflaciently outside the area 

 of greatest concentration to suggest that these plants may be hybrids 

 of B. leucantha x B. sphaerocarpa backcrossed to B. leucantha. The 

 morphological hybrid index suggests that they are Fi hybrids. One of 

 the extra B. leucantha x B. sphaerocarpa hybrids, M-1, lacks a num- 

 ber of major B. leucantha spots and may possibly be a backcross of 

 an Fi hybrid to B. sphaerocarpa. 



