494 JOURNAL Of THE ROYAL HORTICULTURAL SOCIETY. 



Next spring a green leaf is unfolded from the bud and expands 

 above the surface (fig. 115, c). In the autumn of the same year 

 more roots arise, distributed about the apex in the same manner, 

 those on the under side being contractile. The apex is thus 

 continually depressed (fig. 115, d), and from year to year the tuber 

 gradually approaches its normal depth of 10 centimetres. When 

 this is attained, as shown in fig. 115, e, the new roots are no longer 

 appreciably contractile, and the tuber extends itself in the 

 horizontal direction (fig. 115, /). If raised and placed near the 

 surface the roots formed are contractile, and the sinking process 

 is again exhibited. (Cf. fig. 115, g and h.) As a rule, in Arum new 

 effectively contractile roots arise on the tuber in the late autumn. 

 With winter comes a dormant period ; whilst next spring leaves 

 are unfolded and the roots perform the usual nutritive function. 

 On the ripening of the fruit (August) the leaves and roots are 

 lost and another passive phase entered on (fig. 115, / and h). 

 Finally in the late autumn new roots are developed. . 



The above example, in which some, at any rate, of the roots 

 produced at any given time are non- contractile, leads on to the 

 more usual condition in which only a very limited number of all 

 the roots formed have the power of shortening. This condition 

 is well illustrated by the Crocus. From the base of the old corm 

 thin nutritive roots arise, and in early spring foliage and flowers 

 expand from the buds. As time goes on the base of each bud 

 enlarges and forms next year's corm. During the development 

 of these young corms it is a common occurrence for each to give 

 origin to a solitary spindle-shaped root of considerable succulence 

 and possessed of undoubted contractility. These roots arise at 

 the base of the young corms in early spring, and grow past, or 

 occasionally through, the old corm. They have been occasion- 

 ally noted in the literature ; thus Maw, in his " Monograph of 

 the Genus Crocus," alludes to these structures as being of only 

 occasional occurrence upon established plants, though constant 

 in the seedling. He terms them " ephemeral roots," owing to 

 the fact of their subsequent absorption. Though a storage 

 function has been attributed to them, there can be no doubt of 

 their marked contractile properties, as a result of which the 

 young corms producing them are brought down to the level in 

 the soil occupied by the old and exhausted ones. 



Maw's reference to this subject (I.e. p. 18) is of interest in 



