No. 563] STUDIES ON DIFFERENTIAL MORTALITY 



689 



These conditions were most satisfactorily met in the 

 seeds held over from various pedigree experiments made 

 during the last several years. Coupled with the favor- 

 able points of these are some obvious disadvantages, 9 

 which practically are of relatively small weight in view 

 of the fact that it would require several years work to 

 secure a better series. 



It is unnecessary to devote space to the description of 

 these materials, since the key letters used are those 

 employed in previous papers, in which a large amount 

 of quantitative information concerning them may be 

 obtained. 



Altogether thirteen "experiments" were made. That 

 is, a greenhouse or a section of a greenhouse was filled 

 thirteen times. These experiments are numbered A to 

 M, and the letters separated from the pedigree formulae 

 by dashes in the tables refer to them. As a glance at the 

 tables will show, several different series of seeds often 

 went into a single experiment— the capacity of the small 

 greenhouse being about 3,000 and that of the large green- 

 house about 8,000 pots. The specific details of these 

 experiments seem at present irrelevant. 



9 Chief among these is the age of some of the seeds— resulting in very 



The field cultures were grown in 1908, 1909 and 1910. The sand cultures, 



various fie^ experiments, were carried out in the summer of 1912. Any 

 one who takes the ratio of the seeds germinating to those actually planted 

 for the individual samples will be impressed by the very low percentages 

 of germination in these experiments. This is largely attributable to differ- 



grown under different environmental conditions and that they were germi- 

 nated under conditions which could not be maintained the same from ex- 

 periment to experiment. Inability to control temperature and substratum 



Now it is clear that in these experiments it has not been possible to 

 differentiate between the deaths which occurred in the seed envelopes and 



conditions. ^ This problem^ can not be profitably discussed until experiments 



