258 Maternal Characters in Seeds [CH. 



conditions. But we here meet the old difficulty that the 

 distinction between the living types is often sharp, whereas 

 that between the environments is one of degree. We know 

 also that from the same parents a brood of larvae may be 

 raised some of which will feed up fast and emerge as the 

 autumn form while others will feed up slowly and emerge 

 with all the characteristics of the spring form after the 

 winter, and we can form no idea as to the circumstance 

 which determines the distinction. Though I can give no 

 hint as to the rationale of these occurrences which in 

 the cases of the sex of Aphidae &c. have an extreme im- 

 portance to the progress of genetics, I call attention to 

 them as constituting what must be regarded as the group 

 of phenomena least in accord with the conception that the 

 characteristics of living things are an expression of their 

 factorial composition*. 



5. Maternal Characters in Embryos. 

 A. Seeds of Wheat. 



In certain plants a curious and at present quite unin- 

 telligible phenomenon has been met with. Of this four 

 examples are known. The simplest is that seen when a 

 wheat having long glumes and seeds is crossed with a 

 variety having short seeds in short glumes. This case has 

 been studied by Biffen. The F^ seed, resulting from the 

 cross-fertilisation, is unchanged, and resembles the normal 

 seeds of the mother-plant. When this F l seed is sown the 

 plant in due course flowers. Its glumes are of intermediate 

 length (Fig. 34) and the seeds they contain (/%) are all 

 alike of intermediate length (Fig. 35). Apparently there- 

 fore segregation has not taken place. But when these 

 F z seeds are sown, the plants which they become are 

 respectively longs, intermediates (like /\), and shorts, as 

 regards their glumes and seeds. Hence it is clear that 

 segregation had occurred among the gametes produced by 

 /**!, but owing to some control exercised by the mother- 

 plant, all the seeds have a similar length and appearance. 



* It is not impossible that the difficulties which have been met with 

 in the attempt to pursue Mendelian analysis in the double-brooded 

 Lepidoptera may be connected with the phenomena spoken of above. 



