F. KiDD AND V. West 



133 



He further showed that a better yield was obtained from a given 

 weight of heavy seeds than from the same weight of Hght seeds in spite 

 of the fact that nearly twice as many plants were obtained in the latter 

 case, thus emphasising the fact that from the point of view of the agri- 

 culturalist it is advantageous to use large seed. 



Table XII. 



Results obtained froin 188 grams of each kind of seed sown in 



plots 8-5 metres square. 



Weight of grain 



In view of Lehmann's results reference must be made to Johannsen's 

 (43) work with beans. Taking a mixed sample of seeds of a single variety 

 of Phaseolus, he obtained a simple Galton frequency curve for their 

 weights. As a result of sowing the large seeds he found that the plants 

 produced (self-fertilised) showed a shifting of the modal weight in the 

 direction of large-seededness. Similarly, when small seeds were chosen, 

 the modal weight of the crop obtained from them was found to be smaller 

 than that of the original sample. But when the same experiment was 

 repeated with any one family {i.e. the crop raised from a single seed), 

 selection did not alter the modal weight. 



Johannsen's results might have an application in interpreting those 

 of Lehmann: it is to be noted, however, that in Lehmann's experiment 

 the data do not allow one to determine whether there was any increase 

 or decrease in the average seed- weight. It is also to be noted that he 

 remarks that the plants from the smaller seeds were less well-developed 



