170 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



to form the tuber. The whole of the solitary cotyledon, except 

 a short stalk attaching it to the tuber, remains within the seed 

 and decays with it. The short stalk or petiole of the cotyledon 

 remains attached to the tuber for some time in the form of a 

 minute hooked stalk-like process, which ultimately withers and 

 disappears. The first leaf is produced at the top of the tuber, 

 by the side of the hook-like remnant of the cotyledon, and 

 resembles the perfect leaves of the plant in everything except its 

 smaller size. The subsequent leaves are produced in succession 

 one after the other. Mirbel * gives a different account of the 

 process of germination in this genus, and considers what has 

 just been mentioned as the first leaf to be really the cotyledon. 



My own observations of a great number of seeds and seed- 

 lings lead me to consider the view taken by Gaertner and 

 Treviranus as correct, and to say with the last-named observer 

 that the greater part of the cotyledon remains constantly within 

 the seed, and that the first leaf is a subsequent production, of 

 which not a trace exists in the seed itself. The hook-like 

 remnant of the cotyledon Mirbel seems to have mistaken for the 

 rudiment of the second leaf. 



The primary radicle speedily disappears ; to supply its place 

 a number of adventitious root fibrils spring from the base of the 

 tuber. 



On the first glance the vestige of the cotyledon is very likely 

 to fre entirely overlooked, or, indeed, it may have disappeared 

 during the growth of the little plant ; in either case, the first 

 leaf is very likely to be mistaken for the cotyledon — an error 

 which can only be detected by close observation of the seed 

 before and during germination. 



Again, when the first leaf is injured during its growth — as 

 was the case in several specimens that came under my notice — 

 another error may readily be made, viz., that of considering the 

 second leaf as a cotyledon opposite to the first, the difference in 

 size and degree of development being accounted for by the injury 

 or disease that has befallen the supposed fellow-cotyledon. This 

 error, like the former one, is readily corrected by examining a 

 large number of seeds and seedlings in all stages of their 

 development. 



One of the specimens here illustrated (No. 6) has precisely 

 * Ann. du Mus., xvi. 454, t. vi. (xxi.), f. 1. 



