58 THE AxMERICAN BOTANIST 



from the somewhat unlooked for behavior of a stray excep- 

 tion or two of this kind, accidently found that the following 

 peculiarities have been noted. 



I perhaps would not have been any the wiser but for the 

 presence of a large pile of sandy soil that had been lying near 

 an out building since the autumn of 1901. It had been placed 

 there with the intention of grading about the building and was 

 composed enterely of discarded seed tests made at that time. 

 The material was not used as had been expected and the pile 

 remained undisturbed. No additions were ever made to it in 

 the intervening years. On its north side a sturdy honey lo- 

 cust, {Gleditsia triacantlios )t\\enty feet high and ten inches 

 in diameter, marked the spot where an ungerminated seed had 

 been thrown along with other test refuse that accumulated 

 at the time. The pile contained the remains of the average 

 assortment of seeds to be found in a modern seedsman's cat- 

 alogue besides those of many deciduous and coniferous trees, 

 and the spores of cultivated ferns. 



In later years I sometimes used this sandy soil in pot- 

 ting plants and was astonished at the persistence of so many 

 seed coat^ that had undergone little or no change in appear- 

 ance. Eleven years afterwards, (1912) the remains of cu- 

 curbit seed coats were conspicuous so that one could readily 

 identify those belonging to gourds, squashes, pumpkins, and 

 melons. In addition there were others, chiefly those of 

 asparagus, beets, onions, okra, spinach and winter vetches. 

 I also found in my sieve^ at this date, three canna seeds that 

 to all external appearances were uninjured. One I cracked 

 open, finding the seed in apparently normal condition. The 

 other two were placed in a germinator and in a very few 

 days both burst into growth. They were later planted out 

 and developed into large hardy plants; each bearing a large 



