30 



Seed Wheat. 



1 



done with a sickel, the plants being cut close to the ground and as evenly as possible. If 

 the plant had died when a seedling, and before it had produced a stalk, it would not 



enter in to the calculation. Mere 

 dead blades of young seedlings 

 were not counted, and no at- 

 tempt was made to weigh such. 

 There were very few such cases. 

 The counting was done at har- 

 vest time. Uncertain cases were 

 of very rare occurrence. If a 

 plant survived the first few 

 weeks it nearly always reached 

 a growth that was harvested, 

 and if there was no grain on 

 the plant the straw was cut 

 and weighed just the same. 



Few notes were taken on the 

 germination, for the reason 

 that they would have been of 

 rather uncertain value. If no 

 plant appeared at the place 

 where a seed had been sown, 

 that would be no proof that 

 the seed had not germinated, 

 though it would be evidence 

 tending to show that it had 

 not germinated. The plantlet 

 might have been killed b 

 disease or by some insect, 

 considerable number of exami- 

 nations convinced me that most 

 of the cases of this sort were 

 really cases of failure to ger- 

 minate ; but it will be at once 

 seen that any attempt to settle 

 definitely in such a case whether 

 the seed failed to germinate or 

 was killed off by disease would 

 have involved considerable diffi- 

 culty. 



Nevertheless, some attention 

 was given to this feature of 

 the experiment, with the re- 

 sult of showing that so far as 

 observation in the spring could 

 be reasonably carried, it appa- 

 rently showed a lower germina- 

 tion on the part of the smaller 

 seeds certainly a lower plant- 

 producing power. I am in- 

 clined to believe that these 

 observations (see the table 

 on page 31) prove the ger- 

 minating power of the small 

 and shrivelled seeds to be less 

 than that of others ; but I 

 hesitate somewhat in the mat- 

 ter, because of results already 

 published by others tending 

 to show that the germina- 

 ting power of rust-shrivelled 

 seed is actually greater than 

 that of plump seed. As 

 before remarked, these counts 

 of the plants as they ap- 

 peared above ground are not 

 thoroughly reliable evidence as 



Fig. 29. Photograph of three rows of experiment wheat after 

 harvest. The rows from large, medium-sized, and small grain 

 are lettered respectively L, M, and S. The picture shows the 

 relative amounts of straw from the various rows, and the 

 greater proportion of "misses" of the smaller seed. 



