n. PII.OSUS, AND THF.IR NATl'KAI, RKI.ATIONSHIl'. 235 



becomes apparent. D. syJvcstris goes on with new and vigorous 

 developments ; out of the sheath of the petiole is developed the cup, 

 and each pair of leaves is strengthened and rendered rigid by a more 

 perfect system of venation than obtains in D. pilosus. The prickly 

 system is afterwards matured with the same vigour, and the whole 

 gives rise to a very robust plant having flowering heads proportionately 

 much larger than its kinsman. There is still usually an excess of 

 vigour expended in different ways, sometimes in producing very 

 long and foliaceous bracteoles, sometimes in bifurcating a leaf or 

 leaves, and sometimes in producing an additional leaf or pair of 

 leaves at one of the upper nodes. 



In D. pilosus this vigorous growth is pretty well absent through- 

 out. In this species, in place of the cup, there is developed from 

 the sheath of the petiole, which is not greatly expanded, a fringe of 

 bristly hairs, and also some prickles on the earlier nodes of the stem. 

 The cauline leaves, which are the largest of the plant in both species, 

 in D. pilosus are peculiar. They have a naked petiole for the most 

 part, but at the base of the leaf-limb some two or three leaflets are 

 generally developed. The leaf, moreover, has a flabby appearance. 

 It is at this stage that the essential weakness of the plant, as com- 

 pared with D. sy/vestris, becomes apparent. The flower stalks are 

 clothed with weak prickles, and, with the flower-heads, attain a size 

 comparable with D. sy/vestris only when its stem has been cut 

 nearly through so as to allow only two or three bundles of fibres to 

 nourish the plant. 



We have adverted to the not infrequent case of bifurcation (or 

 even addition) of a leaf of D. sylvestris, which, of course, means that 

 the mid-rib of the leaf separates at a certain point into two equal or 

 uiiecjual portions, and these develop proportional independent 

 leaflets. This must be regarded as the most pronounced stage of a 

 l)henomenon which is traceable in both species, but far more 

 frequently in D. sylvestris. It seems reasonable to suppose that the 

 plant exercising this function (of variation) most readily should be 

 the newer form. 



Closely associated with the fibro-vascular tissue giving rise to these 

 variations are the prickles, and some attention should therefore be 

 directed to their structure and distribution. In the advanced stage 

 of D. sy/vestris they almost acquire the consistency of spines, 

 whereas in D. pi/osus they are often represented by hairs. They are 

 all epidermal, and there is no material difference between the hair of 



<) 2 



