Cockayne. — Development of Seedlings. 91 



rising to a height of from 3 cm. to 6 cm., or being even less iu 

 stature, and made up of stout creeping or partially ascending 

 woody branches, bearing erect or semierect narrow cladodes 

 1-4: cm. m length or a little longer, and 1 mm. broad or even 

 mucli less — according to Kirk {loc. cit., p. 108) o^oi^^- ^^ tV"^- 

 broad — and usually quite leafless. " Leaves only found on 

 very young plants, small, orbicular, emarginate, shortly 

 petioled" (Kirk). 



Unlike the other species of Garmicliaelia of which I have 

 treated up to the present, G. enysii does not go through an 

 early leafy stage with a true stem and large orbicular leaves, 

 to be succeeded by cladodes more or less leafy ; on the con- 

 trary, it developes at once this semi-leafy form, which is suc- 

 ceeded finally by a much dwarfer and quite leafless growth. 

 Garmichaelia nana, a closely related species, behaves in a 

 similar manner. A seedling of this latter from seed sown on 

 the 10th December, 1897 (No. 351), is now (October, 1899) 

 furnished with one long sparsely leafy cladode 9 cm. in length, 

 the leaves small and oblong-emarginate. Very young seed- 

 lino-s of G. monroi, another related species, are behaving in a 

 similar manner, and developing erect cladodes with small ob- 

 cordate leaflets. A plant of this latter species from Mount 

 Isabel, Hanmer Plains, had when collected both the ordinary 

 broad short leafless cladode and the tall narrow seedling form 

 of growth. This seedling form, as exhibited in these three 

 species, reminds one of the second stage of development of the 

 tall leafless Garmichaelias ; so much so indeed that it is hard 

 to believe, when for the first time examining a pot of seedlings 

 of C. nana, e.g., that a mistake has not been made at the time 

 of sowing, and that they are not seedlings of G. Jiageiliformis 

 or of one of its allies. This resemblance between the first 

 state of development of the dwarf Garmichaelias on the one 

 hand and of the later and second stage of development of the 

 tall Garmichaelias on the other hand, coupled with the fact 

 that the former do not go through a first form with true stems 

 and large leaves, seems to pomt out that the ''nana section " 

 are descended from tall leafless ancestors, just as these latter 

 may have descended from the leafy Garmicliaelias. Thus we 

 may have already a clue to the phylogenetic development of 

 the species of Garmichaelia, the leafy forms, such as G. grandi- 

 iiora, which require a moist atmosphere, being probably the 

 earliest, and the extremely xerophilous G. monroi, G. nana, and 

 G. enysii forming the most recent link in the chain, with G. 

 Hagelliformis and its allies occupying a position midw^ay. This 

 seems the more likely since in G. U7iifiora there seems to be a 

 connecting-link between G. nana, &c., and G. Jiageiliformis, 

 &c., its final form looking like an arrested early seedling form 

 of the ''nana section" (iig. 18), vfhile G. Jiageiliformis, &c.. 



