fi 



nntliers, but more comnioiilv the style is sliort and stont and does 

 not attain to the anther level; very often the stigma may he sessile- 



The vegetative, organs display a wide range of variation, and 

 the genus includes herbs, subshrubs and shrubs, erect, prostrate 

 or scandeut, whicli may be densely leafy or almost leafless, more 

 or less succulent or sliarplj" spinous. 



The leaves, when woU' developed, are usually linear, linear- 

 lanceolate or acicular, more or less keeled, and are generally- 

 larger towards the base of the plant; the floral bracts are similar 

 to the leaves. T. euphovhioides, with its large -amplexicaul 

 leaves, is (juite a unique specie^s. In many species the leaves are 

 merelv subulate scales and both thev and the bracts may have 

 finilniated or scurf}^ edges or blackisli, acuuiinate tips. 



In the spinous speeie3 tlie leaves may l)e closely folded willi 

 sliarp edges mid form spines [T . si^inosum^ T . spirndo-^uv}), or tliey 

 maj^ be solid terete .S2)ines as in T. pungens. The branches also 

 end in spinoiu tips. 



Tlie stems are usually grooved^ the ridges in many species, as 

 for instance in J, lincatum, being pronounced, and in T. av(j}t- 

 Josum they are extended to form definite wings. The species with 

 Kpinous stems (T. rigidum^ T. IlystrLv, etc.) are peculiar in liaviug 

 an irregular wrinkled surface, and the cortical tissue in these 

 species as in the others with few leaA^es is of cliief importance 

 in assimilation. 



^ 



Pubescence is not common, but it may be found both among 

 the spinous and leafy species. 



In some species tliere is a good deal of variation, and it is not 

 always easy to assign a specimen very definitely to a particular 

 species. This occurs especially witli ^specimens" of T. virgatinn. 



f 



In 



sucli cases, where species occupy the same geographical areas, it 

 seems probable that hybridisation may account for the observed 

 resemblances. 



^ Owiug to the general precision of the work of Sonder and De 

 Candolle it has not lieen found necessary to reduce many of their 

 species, but iu_ one instance considerable confusion has arisen. 

 Tlie type of Linnaeus' species T . paniculatuni, preserved in the 

 Linncan Herbarium, and described in the Mantissa, p. 54, does 

 not appear to liave been seen by any other writers on the genus. 

 Sonder quotes a specimen preserved at Stocklndm as Linnaeus' 

 type, but this is an entirely different plant no-reeing with 

 T. virgat„m Lamarck. Thunberg has two sheets labelled 

 I, strictum J and 7 in his herbarium which hear specimens fS 

 spec dextr.* 7 spec sinlstr.) also referable to T. rhnaium and 

 the large number of specimens collected since, some of which 

 have been referred to T. Drv,,canum, A.DC, and T, ^,.,u- 

 rflm«..m A.DC all belong to 7'. ^iraafum, Lam.. whi<-,h is Ihe 

 T pamcvlatum of Sonder. De Candoflo has lal,elk'd some of he 



specimens which he afterwards placed under 7', rn^ucuh'f.T 

 Send., with the name T. cucuUatum (DC. Prodr. iv. C5G) bul 



erect i~ 



lerjT 



