A Branch of the Lamarckiana- family. 261 



from these the beds contained nothing but pure 0. La- 

 marckiana: I examined them repeatedly and thoroughly 

 during the whole time that they were in flower. 



I harvested the seed in October and took care that as 

 few as possible fell to the ground ; for it still contained 

 a good deal of seed from the original sowing which I 

 expected to germinate in the following spring. 



I had already become acquainted with this belated 

 germination in one particular case. In March 1887 I 

 sowed some thousands of seeds ; they germinated during 

 the course of the whole summer and at the end of each 

 month the seedlings were counted and pulled up. Up to 

 the middle of April 908 germinated ; between then and 

 the middle of May 288. From the 14th of May to the 

 14th of July only 64 germinated: from then till the 14th 

 of September 130; and between that date and the middle 

 of October only 6. During the winter there was no ger- 

 mination although the sowing-pans were protected from 

 the frost and the circumstances were favorable in every 

 other way. Up till the middle of March 1888 there ap- 

 peared only three seedlings. But then — in the second 

 spring — an extraordinarily large number appeared.-^ 

 Within 14 days 272 plants had unfolded their leaves, and 

 others followed as before in gradually decreasing num- 

 bers. The latest seeds stayed for 2 or more years in the 

 ground before they germinated. 



In the spring of 1896 therefore I expected the seeds 

 which had remained in the ground from the previous 

 spring to germinate. Whilst this was taking place I 

 counted the mutants and transplanted them singly into 

 pots as before ; I weeded out the normal individuals as 



^ A similar phenomenon is known to occur in the case of species 

 of clover, Primula and many other plants. 



