Index Frequency Method 525 



using powdered malt extract sterilized by filtration in place of the 

 coconut milk as a source of the "embryo factor." This method 

 has also been applied successfully to hybrids between 7ns pseud- 

 acorus and 7. versicolor and between different species of Lilium. 

 In Chapter 26 we called attention to the work of Brink and 

 Cooper on somatoplastic sterility. They have shown that the 

 dominant tissue in a juvenile angiosperm seed is the endosperm. 

 They have further maintained that the collapse of a seed pro- 

 duced from a cross between two species that are wide apart 

 is the result of disharmony in the endosperm rather than in the 

 embryo. A cross which they made between the squirrel-tail 

 grass, Hordeum juhatum, and rye, Secale cereale, illustrates this. 

 Hybrids are readily obtained and the hybrid seed develops 6 to 

 13 days and then collapses. Brink, Cooper, and Ausherman dis- 

 sected the embryos from 9- to 12-day seeds and grew them 

 upon artificial media. Of 81 treated embryos, 34 were free of 

 fungal or bacterial contamination and made considerable growth 

 in the nutrient media. One embryo differentiated normally and 

 was potted up and raised to sexual maturity. A study of meiotic 

 behavior indicated that there was little homology between the 

 parental genomes. By embryo culture somatoplastic isolation 

 may be overcome ' and hybrids produced between species that 

 would never produce viable hybrids in the wild. 



Index Frequency Method 



A method which attempts to picture quantitatively the quali- 

 tative variation exhibited by hybrid swarms has been developed 

 by Anderson largely for Tradescantia hybrids, and has been 

 used by others in studying other genera. This method, criticized 

 by some investigators, is highly subjective in the selection of 

 characters and the way in which it weights them, but it has cer- 

 tain advantages that, provided its limitations are understood, 

 make it a very useful technique for comparing populations. 



Two species are compared with respect to a number of char- 

 acters, each markedly different in the two types. The value is 

 assigned to each character of one species and another value as 

 2, 3, 4, or some other relatively low number to each correspond- 

 ing character in the other species. If the two values are and 2 

 for a certain trait, an intermediate type would be given the value 

 1. If the values are and 4, the intermediate would be 2 and 



