126 Glume Length in Triticum Polonicum 



be classed as smooth. With the object of investigating this, in 1912 a 

 cross was made between a variety of T. polonicum, with an average glume 

 length of 29 mm. and very faintly pubescent, and a variety of T. durum, 

 which will be referred to in this paper as Kubanka. This last is a 

 smooth and otherwise typical example of T. durum, with an average 

 glume length of 12 mm. The first generation was a hybrid of inter- 

 mediate length — actually of an average of 18 or 19 mm. It was, how- 

 ever, remarkable in that it was distinctly pubescent — very much more 

 so than the polonicum parent. The second generation was surprising, 

 for it was soon observed that there was a proportion of plants bearing 

 fully pubescent ears— pubescence, be it remembered, is a dominant 

 character — yet the variety of Polish wheat used as a parent would have 

 been classed as smooth in comparison with such a wheat as Rivet or 

 Essex Rough Chaff. 



At harvest time, a middle glume in the ear of each plant was mea- 

 sured and a curve plotted of the number of plants of each glume length 

 in millimetres (Charts Fig. la). The plants were also classified into 

 pubescent, intermediate and smooth, by means of a hand lens. The 

 polonicum parent would have fallen among the intermediates in this 

 classification, while the smooths were, as far as could be seen, perfectly 

 glabrous. The numbers observed, considering glume length alone, were 

 as follows : 



Expectation 



A glance at the curve of this family will show that there is no dividing 

 line between the longs and the intermediates but, actually, the shorts 

 can be distinguished by eye-^that is to say, an extra short glumed 

 heterozygote. A starved plant, for example, which might fall in the 

 14 or 15 mm. lengths class, has an indefinable something about it which 

 points to its really belonging to the heterozygote class. 



The carrying of large numbers of plants into the F^ generation 

 showed that, while only two mistakes were made at the short end of 

 the curves, it had been impossible to pick out any but the extreme 

 longs with the certainty that they would be pure to their particular 

 length. Considering those with a glume length varying between 10 

 and 14 mm. as being pure short segregates and those between 15 and 

 31 mm. as including both the longs and the heterozygotes, a count of 

 the proportion of pubescent individuals shows that in the short glumed 

 ^ See pp. 130, 131 and explanation on p. 133. 



