The total number of species of vascular plants, together with such varieties as are of equal rank to many admitted species, is 1843 of which 166 are pteridophytes {Filices 147), 20 gymnosperms and 1657 angiosperms (monocotyledons 428 and dicotyledons 1229), and they belong to 109 families and 383 genera. The average number of species to each family is 16.9 and 4.8 to each genus.

The largest families and genera are as follows: — (families) Compositae 256 (compound 55), Filices 147 (comp. 13), Cyperaceae 133 (comp. 15), Gramineae 131 (comp.21), Umbelliferae 89 (comp.7), Orchidaceae 71 (comp. 1), Ranunculaceae 61 (comp. 13). Rubiaceae 55 (comp. 12), Onagraceae 45 (comp. 8), Epacridaceae 44 (comp. 7), Leguminosae 38 (comp. 5), Borraginaceae 33 (comp. 4), Rosaceae 29 (comp. 9), Cruciferae 27 (comp. 2), Gentianaceae 24 (comp. 5), Pittosporaceae 22 (comp. 4), Halorrhagaceae 22 (comp. 3), Caryophyllaceae 21 (comp. 3), Myrtaceae 20 (comp. 2), Thymelaeaceae 20 page 401(comp. 6), Araliaceae 20 (comp. 4); (genera) Hebe 66, Carex 59, Celmisia 54, Coprosma 48, Ranunculus 47, Epilobium 41, Olearia 35, Senecio 35, Poa 33, Myosotis 32, Aciphylla 29, Dracophyllum 28, Carmichaelia 25, Cotula 25, Pittosporum 22, Anisotome 22, Gentiana 22, Raoulia 22, Uncinia 20, Pimelea 20, Hymenophyllum 19, Danthonia 17, Asplenium 15, Juncus 15, Acaena 15, Blechnum 14, Scirpus 14, Thelymitra 14, Veronica 14, Pterostylis 13 and Euphrasia 13.

Although the above figures show the relative importance of the larger families and genera, so far as the flora is concerned, they exaggerate the part that some play in the vegetation. In this regard many small genera are of more moment although they may be of restricted distribution; the following to which the number of species are appended are examples: — Cyathea 5, Hemitelia 1, Dicksonia S, Pteridium 1, Gleichenia 6 (Filices); Agathis 1, Libocedrus 2 (Cupressac); Typha 1 (Typhac); Freycinetia 1 (Pandanac); Arundo 2 (Gramin.); Mariscus 1, Schoenus 7; Gahnia8 (Cyperac); Rhopalostylis 2 (Palmae): Leptocarpus 1, Hypolaena 1 (Restionac); Rhipogonum 1, Cordyline 4, Astelia 12, Phormium 2, Chrysobaczron 3 (Liliac); Nothofagus 5 (Fagac); Elatostema 1 (Urticac); Elytranthe 4 (Loranthac), Muehlenbeckia 6 (Polygonac), Carpodetus 1 (Saxifrag.), Weinmannia 2 (Cunoniac), Rubus 5 (Rosac), Edwardsia 4 (Legum.), Coriaria 4 (Coriariac), Pomaderris 5 and Discaria 1 (Rhamnac.), Aristotelia 2 and Elaeocarpus 2 (Elaeocarp.), Plagianthus 2 and Hoheria 5 (Malvac), Melicytus 5 (Violac), Leptospermum 4 and Myrtus 3 (Myrtac), Fuchsia 3 or 4 (Onagrac), Nothopanax 7, Pseudopanax 7 and Schefflera 1 (Araliac); Griselinia 2 (Cornac); Gaultheria 5 (Ericac.); Cyathodes 5, Leucopogon 3 and Epacris 3 (Epacrid.); Suttonia 8 (Myrsinac); Olea 4 (Oleac); Geniostoma 1 (Loganiac); Parsonsia 2 (Apocynac); Vitex 1, Avicennia (Verbenac); Myoporum (Myoporac); Nertera 6 (Rubiac.);. Alseuosmia 4 (Caprifoliac.); Selliera 1 (Goodeniac.); Pleurophyllum 4, Haastia 4, Leucogenes 2, Cassinia 6 and Brachyglottis 2 (Compositae).

In contradistinction to the above, the following genera are so rare that their absence from the vegetation would not be noticed: — Simplicia 1, Amphibromus 1 (Gramin.); Hydatella 1 (Centrolepidac.); Iphigenia 1 (Liliac); Hypoxis 1 (Amaryll.), Bagnisia 1 (Burmanniac), Caleana 1 (Orchid.), Phrygilanthus 2 (Loranth.), Logania 1 (Loganiac.) and Tetrachondra 1 (Tetrachondrac).

Hybrids.

Up to the year 1912 5 hybrids had been recorded for the flora though with some diffidence (Cockayne, L. 1912a: 30–31), but in this paper it was suggested that variation in the highly-variable genera Hebe, Celmisia and Acaena was possibly due in large measure to hybridism. Five years later, the study of wild hybrids began in earnest, so that, at the present time, no less than 290 groups of such have been noted, a large majority consisting not merely of a few individuals more or less similar but of great polymorphic swarms.

page 402

The 290 groups of hybrids belong to 42 families (40% of the families) and 92 genera (24% of the genera), the largest of which are as follows:— Compositae 64 groups of hybrids, Scrophulariaceae 36, Filices 21, Gramineae 17, Acaena and Coprosma 14 each, Cyperaceae 12, Hebe 31, Celmisia 18, Coprosma 14, Olearia 13, Acaena 10, Asplenium 9 and Ranunculus 7. In nearly all the above genera additional hybrid groups are certainly expected. In addition, very large polymorphic swarms occur in the following genera:— Danthonia, Uncinia, Luzula, Phormium, Nothofagus, Mida, Pittosporum, Melicope, Coriaria, Aristotelia, Myrtus, Epilobium, Nothopanax, Fuchsia, Gaultheria, Apium, Corokia, Dracophyllum, Parsonsia, Veronica, Alseuosmia, Gnaphalium, Helichrysum, Cassinia, Craspedia, Cotula and Senecio.

Occasionally, hybrids occur in which 3 species are concerned, e. g, in Cassinia, Coprosma, Coriaria and Alseuosmia. Also intergeneric hybrids are not unknown, e. g. Ewartia × Helichrysum, Nothopanax × Pseudopanax, Helichrysum × Gnaphalium, Leucogenes × Raoulia and possibly by Helichrysum (subgen. Ozothamnus).

Generally, the species which cross are of the same life-form, but this is not always the case. Thus the tree-form crosses with the divaricatingshrub form, e. g. Plagianthus betulinus × divaricatus; the tall bushy-shrub with the divaricating-shrub, e. g. Corokia buddleoides × Cotoneaster: the low canopy-tree with the liane, e. g. Fuchsia excorticata × perscandens; the shrub by the slender slightly woody mat-plant, e. g. Helichrysum depressum × bellidioides: the liane by the mat-shrub, e. g. Muehlenbeckia complexa × ephedroides; the tree by the summergreen semi-woody plant, e. g. Coriaria arborea × sarmentosa; the cupressoid-form by the leafy shrub, e. g. Hebe Astoni × buxifolia; the dense cushion-plant by the erect low semi-woody plant, e. g. Raoulia bryoides × Leucogenes grandiceps.

With regard to the validity of the 290 groups of hybrids there is no doubt as to the hybrid nature of 230 of them, most of these having been studied more or less intensively in the field and a good many have been brought into cultivation, especially by W. A. Thomson, J. S. Thomson and G. Simpson of Dunedin, H. H. Allan and myself. Moreover Allan has already synthesised two of the hybrids, and seed sown from others has produced polymorphic progeny. The hybrid groups range from those whose parents usually grow in close proximity to those the parents of which very rarely meet. In certain cases, the unpremeditated action of man with his burning &c. has led to great increase in the number of hybrid individuals, e. g. in Acaena, Myrtus, Hebe, Celmisia and Cassinia. In the case of Acaena, hybrids occur between the exotic A. ovina and the endemic A. inermis, A. Sanguisorbae var. pusilla and A. novae-zelandiae.

So far as observation and experiment go, it seems safe to assume that most of the hybrids produce viable seed, in fact, so far the only more or less sterile hybrid is one I described as a species many years ago under the page 403name Rubus Barkeri. This plant — only the one was found wild — is readily reproduced vegetatively so that it is quite common in cultivation, yet only twice has it been known to bloom since its discovery some 30 years ago. Rubus parvus — a creeping and rooting mat-plant — is one of the parents and probably R. australis (liane) the other parent. Recently Dr, W Mc. Kay of Greymouth (NW.) has discovered a most closely-related plant which produced a fair amount of fruit on its shaded shoots, but none where exposed to bright light. Most likely both plants are of the F 1. generation and those of the F 2. &c. generation must be exceedingly rare.

From what has gone before it will be expected that from time to time hybrids have been described as species. This expectation is realised in the fact that no less than 42 hybrids masquerade as good species in the Manual of the New Zealand Flora. Also the "variability" of no few "variable species" is due entirely to crossing constantly taking place between the jordanons of which they are composed. This has apparently happened to such an extent in certain species that the jordanons have been entirely swamped out by the hybrids and such alone remain. Over wide areas this seems to be the case with Leptospermum scoparium and Phormium tenax. Near Queenstown (SO.) swamping can be seen in progress in a small piece of Nothofagus forest where hybrids of the × N. cliffusca group are now in greater numbers than the parent species, also in indigenous-induced montane Cassinia shrubland it is often difficult, if not impossible, to pick out the species, though these are constant when one alone is present. In the case of the endemic genus Alseuosmia, polymorphy reaches its highest degree. Theoretically, the genus embraces 4 (Hooker, Kirk, Cheeseman) to 8 species (A. Cunningham), but where all these grow together a polymorphy unspeakable occurs, perhaps impossible to disentangle. Taking the numerous polymorphic groups ("variable species") it should be now clear enough whence comes the reputed, "extreme variability" of the New Zealand flora, yet, in the ordinary taxonomic sense, if the epharmones be excluded, there is no variability whatsoever, but merely a polymorphy due to the grouping together as a species two or more jordanons (closely allied or otherwise) and the hybrids between them.