AV. O. Backhouse 133 



The cause of the suppression of colour in this particiilar cross must 

 be sought for in the shape of an inhibitor, brought in, either by Polish 

 wheat and meeting something in Rivet to release it, as it were, or vice 

 versa ; for this particular strain of polonicum crossed with coloured 

 varieties other than Rivet gives coloured descendants, in climatic con- 

 ditions under which, crossed with Rivet, they are colourless. 



EXPLANATION OF CHART (on pp. 130, 131). 



Fig. la. Family 74/14. Kubankax Polvnicum. 



Curve of glume length plotted for 227 iudividuals, showing that, whereas the short 



glumed individuals, namely those whose glume length varies between 9 and 14 mm., 



are easily distinguished from the rest, there is no discontinuity in the curve between 



the heterozygotes (of, say 15 mm. to 22 ram.) and the bulk of the homozygous longs. 

 Fig. lb. Curve of those individuals which were completely felted. It will be seen that 



there are few of heterozygote glume length and no longs. 

 Fig. 2. Family 104/16. Polonicum x Turgidum 179. 



The fully coloured individuals. These, with one single exception, are among the 



short glumed plants. 

 Fig. 3. Those individuals which were not fully coloured and classed as tinged. Owing to 



gradation in colour from plainly tinged to faintly tinged or stained white, this curve 



is only of relative value, but serves to show that among the real longs there were only 



three or four with a trace of colour. 

 Fig. 4. The same curve as in Fig. .5, resolved into felted individuals — shown by the whole 



line — and those classified, without the aid of a hand lens, as smooth — shown by the 



dotted line. 

 Fig. 5. Curve of the whole family of 692 plants, plotted for glume length alone. 



