220 PROCEEDINGS OF SECTION D. 



A somewhat different case to any of the preceding is that of 

 Parsonsia (Apocynaceae). The two New Zealand species of this 

 genus, as I have already shown,^ undergo a remarkable series of 

 changes in leaf form during their ontogeny, but these cannot be 

 referred to the effect of the environment ; they are hereditary 

 merely, and not epharmonic at the present time. In the liane, P. 

 heterophylla, I have seen a reversion-shoot to bloom freely, though 

 this is probably not a common occurrence. In P. capsularis, on the 

 contrary, the very narrow-leaved intermediate stage frequently 

 blooms, and in a montane variety this stage remains fixed, this 

 variety or species extending over a considerable area and growing 

 both under xerophytic and mesophytic conditions. The case of 

 the two species of Weinmannia (Cunoniaceae), is somewhat similar 

 to that of Parsonsia in so far as the juvenile form is not epharmonic. 

 Both species {W. racemosa, W. sylvicola) show a reduction in leaf- 

 form in the adult stage, that of the former being simple, but the 

 latter being always more or less compound ; in fact, since the 

 flow^ers, etc., are virtually the same in both species, W. sylvicola 

 may be considered a fixed juvenile form of W. racemosa. Further, 

 both species can remain at an earlier juvenile stage as shrubs with 

 compound leaves, and bloom abundantly, but in this case their 

 station is not the rain-forest, but the more xerophytic heath. 



The case of Dacrydium intermedium (Taxacese) is of consider- 

 able interest, and recalls that of the xerophytic whipcord veronicas. 

 This tree, like so many of the New Zealand taxads, has a juvenile 

 form very distinct from that of the adult. It has drooping branches 

 which bear long, narrow, spreading pointed leaves. These pass by 

 gradual gradations into the short, thick, rhomboid, imbricating 

 reduced leaves of the adult, which are closely pressed to the branch. 

 In Stewart Island^ the adult form is not always reached, and 

 juvenile flowering and fruiting trees are quite common, and form 

 the variety gracilis of T. Kirk.^ 



The so-called variety serratum of the mesophytic small forest- 

 tree, A/'o//io/)a«a;c Edgerleyi {ArsiViRceie) is evidently a fixed and flower- 

 ing juvenile form of the latter, and its leaves are more hygrophytic, 

 a condition of affairs which may be referred perhaps to the relation 

 of the seedling to the moist forest floor. The variety trifoliatnm of 

 Pseudopanax crassifoUum (Araliaceae) may also perhaps be con- 

 sidered a fixed earlier stage of development. 



The herbaceous plants do not furnish nearly so many cases of 

 discontinuous stages of development as do the ligneous ; two, how- 

 ever, seem worthy of mention. The celebrated mountain lily 

 {Ramtnculus Lvallii), distinguished from all other members of the 

 genus but one by its peltate leaves, has seedling reniform leaves, the 

 peltate form arising, I am virtually in a position to demonstrate, 

 through concrescence of the basal lobe^ Be this as it may, occa- 

 sional adult flowering plants are encountered in which peltate 



1. Rep. A.A.A.S. 11; 486, 1908. 



2. Cockayne, L. : " Report on a Botanical Survey of Stewart Island," p. 17, 1909. 



3. " Forest Flora," p. 224, 1889. 



