1 '1'2 Phyfiiologiral Frc-detvnin nation 



size of the grains. Harris {i.e.) found in the case of Phaseolus vulgaris 

 that the weight of the seeds increased in relation to the percentage of 

 ovules which failed to develop. 



Von Lochow(i8), working with rye plants of the same origin, showed 

 that the average grain-weight was higher for the plants with spikes 

 incompletely filled with grains than for those with spikes completely 

 filled with grains, the proportions being 39 : 32-5; but that the total 

 yield of grain was considerably in favour of the latter, the proportions 

 ill this case being 72-5 : 90. As a result of his experience Von Lochow 

 nuido the interesting observation that heavier seed of unknown ancestry 

 might be inferior to smaller seed of the same origiii. 



Owen (50), as a result of five seasons' work with beans, found that by 

 limiting the yield of the plant to one pod the weight of the seed was 

 greatly increased, so that the question still to be answered is whether it 

 is not the number of seeds developing that controls their average size 

 even in such experiments as those of Combes quoted above. 



iyh) Deler mined by the PosUion of the Seed in the Fruit, of the Fruit 

 on the Inflorescence, and of the Inflorescence on the Plant. 



Where the size of the seed is obviously mainly controlled by its posi- 

 tion on the parent plant, the presumption is that differences obtained 

 by the use of seeds of different size are to be accounted for by physio- 

 logical pre-determination. 



Several recent authors have worked out on a large scale correlations 

 between seed-size and the position of the seed in the fruit and on the 

 inflorescence, and of the inflorescence on the plant. Halsted and 

 Owen (28) found that the heaviest seeds of Scarlet Kunner beans occur 

 in the three-seeded pods and in the middle position in these pods; two- 

 seeded pods and seeds from the basal position give the smallest average 

 weight, nevertheless seeds from the two-seeded pods show the greatest 

 average germination percentage and the greatest average weight of 

 seedling per plants This result points to a correlation between vigorous 

 growth and seeds of small size, which would appear to be one of physio- 

 logical pre-determination, but unfortunately these experiments were 

 not carried far enough to enable one to decide this point with certainty. 



1 Cuniminf^s (12) observed that in beans 4!) per cent, of the small seed were found to 

 occur in the basal end, while 18 per cent, occurred in the middle of the pod. In the case 

 of garden peas the small seeds were almost always found at the ends of the pods, with one 

 end as prolific as the other. 



