96 University of California Publications in Botany [Vol. 5 



pressing the weights of the seeds in terms of 100 individuals, a 

 true average weight per seed could be calculated. For the same 

 reason it was felt that few individuals, if any, weighed much 

 more or less than this average or, in other words, that every one 

 of the hundred seeds of group 9, for example, which weighed 

 0.0065 grams, corresponded, more or less closely, to a theoretical 

 weight of 0.000065 grams. The above is also, in effect, a com- 

 bined table expressing the various sizes and densities of the 

 different groups of seeds, as well as their weight, since all three 

 characters were found to be so closely correlated as^ mutually to 

 represent one another. Groups 1 and 2, the ill-formed seed, 

 showed naturally a considerable difference in weight among 

 the groups of ten seeds, since they included all abnormally 

 shaped seed, from those that were simply shrunken and wrinkled 

 to those whose seed-coats were flattened and practically empty. 

 The minute seed, that was picked out in the final examination 

 after grading from the group of least density, is included in 3. 



IV. GERMINATION OF THE SEED 



1. In the Germinating Case 



For germinating the seeds it was necessary to use an im- 

 provised germinating case. This was made by heavily insulating 

 a large glass-sided box with herbarium blotters and using a 

 16-candle power electric light, immersed in water, to obtain 

 the desired temperature within the box. It was so arranged that 

 the germinating case could be entirely closed after the seed had 

 been put in and the water about the electric light renewed 

 through a tube. The jar of water in which the light was placed 

 was tightly surrounded by a blotter and the opening at the top 

 of the jar was covered with a zinc plate so that practically 

 no light could be seen within the box. The blotting-paper in- 

 sulation was left free on one side and could be removed when 

 observations of the progress of germination were to be taken. 

 The seeds were counted out from the packets in which they had 

 been placed after weighing onto circular pieces of blotting-paper 



