238 
weeks after the root-tendril grew into the tube, a nutritive 
root developed, in the usual position for such, between the leaf 
base and the root-tendril, but applied itself to the tube, cur- 
ved around it, and ceased growing after attaining a length of 
15 centimetres, the two roots hence reversing their respective 
physiological functions and properties. In other cases, in which 
the nutritive root appeared soon after the root-tendril grew 
into the darkened tube, the former retained its normal proper- 
ties and two nutritive roots were formed at the some node but 
no root-tendril. 
There appears to be no reason for not regarding the root- 
tendril of Vanilla aromatica as a true but not very highly spe- 
cialized or differentiated tendril, which the aerial habit of the 
roots has allowed to become a useful evolutionary development. 
That an aerial root may become changed into a true tendril 
is not surprising when we remember that tendrils may be 
modified stems, branches, floral peduncles, leaves, leaflets, 
petioles, stipules or pulvini. 
Conclusion. 
In the preceding pages it has been shewn, that a series of 
connecting forms exist between hooks such as those of Uncaria 
and the highly specialized tendril of the Passiflora type. The 
connections are however purely physiological, the various clasping 
organs arising not from one but from several morphological 
members, so that in phylogenetic origin the different morpho- 
logical series of clasping organs follow more or less parallel 
lines of increasing physiological development and differentiation. 
In terms of physiological differentiation alone, the different 
plants may be arranged in a nearly linear arrangement of 
ascending physiological complexity. Commencing with hooks 
such as those of Caesalpinia, Rubus fruticosus. Acacia etc. in which 
the hooks are accidentally of use in climbing or for fixing 
the plant, but do not thicken when in contact , we readily 
proceed to Luvunga, which possesses non-irritable spines and 
