120 GROWTH OF PLANTS 



(50° F) gave 35 and 32 per cent of root growth and no epicotyl growth, 

 because there was not a low-temperature period to after-ripen the epicotyls 

 after the roots had grown. Judging from the later figures in this table, 

 the six months' initial period in the greenhouse mterfered with the after- 

 ripening for root growth. Compare 35 and 32 per cent with the percent- 

 ages below in the same columns (Table 15). Thi-ee months at 5° C (41° F), 

 followed by three months in the greenhouse and then three months at 5° C 

 (41° F), gave 83 and 84 per cent of root production and 54 and 38 per 

 cent of shoot production. When the second cold period was lengthened 

 to five months, there was about the same root production but much higher 

 shoot production — 80 and 80 per cent against 54 and 38 per cent. As 

 Table 15 shows, 10° C (50° F) is almost as effective as 5° C (41° F) for 

 after-ripening for both root and shoot growth. 



Table 16 lists several other seeds that belong to this class — Caulo- 

 phyllum thalidroides, Smilacina racemosa, and Trillium erectum strictly 

 so because there is very little root production ^\'ithout a prechilling period. 

 Polygonatum commutatum, Sanguinaria canadensis, and Convallaria majalis 

 belong in part to this class of seeds and in part to the previous class, for 

 there is considerable root production ^\ithout a prechilling period; but 

 such a period increases considerably both the percentage and speed of 

 root production. For instance, Convallaria seeds give 46 per cent root 

 production after several months in the soil without prechilling, whereas 

 seeds that have been stratified for three months at 5° C (41° F) give 

 92 per cent root production rather promptly. If one examines the length 

 of time each of the low-temperature periods (and also the high-temperature 

 periods) requires according to the optimum conditions Barton has found 

 for getting the epicotyl ready to grow into a plant, one will realize not 

 only the complexity of the process but its time-consuming nature. This 

 adds up to 8.5 months for Caulophrjllum seeds, 7 to 18 months for Con- 

 vallaria, 11 months for Polygonatum, 12 months for Sanguinaria, 14 to 

 19 months for Smilacina, 9 to 14 months for Trillium erectum, and 9 to 

 12 months for T. grandiflorum. So far as persistence of the species is con- 

 cerned, one wonders whether nature has not overdone dormancy in these 

 seeds, certainly as to complexity and perhaps as to time required. Yet if 

 nature wanted to become even more complex and difficult she could do so 

 right in line with the things we have noticed in the several classes of dor- 

 mancy mentioned above. Seeds that required initially a high-temperature 

 period in the soil to overcome coat resistance, followed in succession by a 

 low-temperature period to after-ripen for root growth, a high-temperature 

 period for root growth, a low-temperature period for epicotyl after-ripening, 

 and a high-temperature period for growth of the plant would add one more 

 step to the after-ripening process. If such a temperate zone seed existed 

 and matured in the late fall, it would not germinate until the third spring 

 after maturing: it would be a three-S^ear seed. Has nature gone to this 

 extreme with any seed? If so, the folly has not been discovered. 



