Seeds were taken from one plant of either species and from two plants of the suspected hybrids. In the latter case 9 seedlings of plant A. and only 7 seedlings of plant B. survived the seedling stage. These were eventually transferred to the garden area. A second set of 19 viable seedlings was obtained from plant A. in the second year, which, allowing for deaths of weak seedlings, represented an effective germination rate of only 34%. These seedlings were kept in the glasshouse.

An account follows of those characters of the 2 species in the area which were found to be of value when investigating the hybrids.

Acaena novae-zelandiae.

Germination rate, 100%; cotyledon laminae averaging 5 mm long; first leaf trifoliate; growth habit of mature plants trailing with few lateral branches; average diameter of stems at 7th internode 2.25 mm; stems clothed with unicellular, thick-walled hairs arising from pronounced multicellular bases; mature stems red in colour; adult leaves with 5 pairs of leaflets and a terminal one under glasshouse conditions, or 6 pairs of leaflets and a terminal one in the garden; average length of the terminal leaflet 11 mm; length ratio of one leaflet of the second pair from

Text-fig. 4.—Frequency distributions of hybrid indices, Prince of Wales Park. 1, Garden plants. 2, Glasshouse plants. N.Z. = Acaena novae-zelandiae. H. = Suspected hybrid progeny A. = Acaena anserinifolia.

the tip / one leaflet of the third pair is 10/7.6; upper surface of leaflets green only, lower surface of leaflets green only, hydathodes at serration tips white in glasshouse plants, crimson in garden plants; upper surfaces of leaflets glabrous; calyx lobes green only; average length of longest fruit spine 9.8 mm; fruit spines bright crimson; pollen fertility 98%; (for method of determination see Owczarzak, 1952) chromosome number n = 21.

Acaena anserinifolia.

Germination rate 96%; average length cotyledon laminae 2.4 mm; first leaf simple; growth habit of mature plants spreading with many lateral branches; stems clothed with unicellular hairs lacking multi-cellular bases; mature stems olivebrown

in colour; adult leaves with four pairs of leaflets and a terminal one under glasshouse conditions or 5 pairs of leaflets and a terminal one in the garden; length ratio of one leaflet of the second pair from the tip/one leaflet of the third pair is 10: 4.7; upper surface of leaflets with a red-brown colouration at serration tips and along margins; lower surface of leaflets with pink-purple veins; hydathodes pink; upper surfaces of lower leaflets pilose; calyx lobes reddish on the abaxial surface; average length of longest fruit spine 5.2 mm; fruit spines pale-brown to reddish-brown; pollen fertility 94%; chromosome number n=21.

The above contrasting characters of the two species were used according to Anderson's hybrid index method (Anderson, 1949) for the investigation of the suspected hybrids. In each case the Acaena novae-zelandiae character was rated as 0 and the A. anserinifolia character as 1, 2 or 3 depending upon the number of intermediate grades.

(a) Length of cotyledon: 4.5–5.5 mm. = 0 (A. novae-zelandiae); 3.5–4.5 mm. = 1; 2.5–3.5 mm. = 2; 2–2.5 mm. = 3 (A. anserinifolia).

(b) First leaf form: Trifoliate = 0 (A. novae-zelandiae); simple, deeply incised = 1; simple = 2 (A. ansernifolia).

(c) Growth habit: For each plant the total number of visible nodes possessed by lateral branches was determined. 0–10 nodes = (A. novae-zelandiae); 10–20 = 1; 20–40 nodes = 2; more than 40 nodes = 3 (A. anserinifolia).

(d) Stem colour: Red = 0 (A. novae-zelandiae); olive-brown = 1 (A. anserinilolia)

(e) Hair bases: Pronounced multicellular bases = 0 (A. novae-zelandiae); small multicellular bases = 1; no multicellular bases = 2 (A. anserinifolia).

(f) Number of leaflet pairs: 5 pairs (or 6 in the open) = 0 (A. novaezelandiae); 4 pairs (or 5 in the open) = 1 (A. anserinifolia).

(g) Leaflet size gradation: Ratio more than 10.7 = 0 (A. novae-zelandiae); ratio 10.5–10.7 = 1, ratio less than 10.5 = 2 (A. anserinifolia).

(h) Colour of leaflet upper surfaces: No red colouration along upper margins of leaflets = 0 (A. novae-zelandiae); red-brown colouration along margins = 1 (A. anserinifolia)

(i) Colour of leaflet under-surfaces: No purple colouration = 0 (A. novaezelandiae); some degree of purple colouration on veins and surface between = 1 (A. anserinifolia)

(j) Hydathode colour: Hydathode white (or crimson on plants in the open) = 0 (A. novae-zelandiae); hydathode pink = 1 (A. anserinifolia)

(k) Upper surface hairs on lower leaflets: Lower leaflet pair glabrous on upper surface = 0 (A. novae-zelandiae) lower leaflet pair moderately hairy on upper surface = 1 (A. anserinifolia).

(l) Colour of calyx lobes: Green only = 0 (A. novae-zelandiae); reddish colouration on margins and midrib of abaxial surface = 1; general reddish colouration over abaxial surface = 2 (A. anserinifolia)

(m) Colour of spines: Distinct red colouration = 0 (A. novae-zelandiae) brown colouration = 1; (A. anserinifolia.)

(n) Length of longest spine: 9–11 mm = 0 (A. novae-zelandiae); 7.5–9 mm = 1; 6–7.5 mm = 2, 4–6 mm = 3 (A. anserinifolia)

Suspected hybrids in the field.

The plants forming this colony were inspected fairly closely on several occasions and a comparison of specimens from seven well separated points revealed no significant variations in vegetative or reproductive features. There are, therefore, three possibilities: (a) that these plants are first generation hybrids; (b) that they are a clone derived from a single F₁ hybrid; (c) that they are a clone derived from an F₂ or back cross hybrid. The following data support the first and second possibilities.

As no seedling stages were available these plants could be scored for the following characters only:—

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

For these characters Acaena novae-zelandiae scored 0 and A. anserinifolia 14 so the suspected hybrids were exactly intermediate.

Pollen slides were made from four inflorescences possibly representing four different plants. The percentages of fertile pollen were 63%, 0%, 18%, 48%.

Text-fig. 5.—Population pattern, Lyall Bay-Seatoun.

Progeny of suspected hybrids.

These were chiefly studied in the vegetative state. The 16 surviving plants from the first sowing were planted out in the garden and the 19 seedlings from the second sowing were kept in the glasshouse. Only 6 of the garden plants came to flower, and their pollen fertility percentages were: 54%, 81%, 80%, 79%, 65%, 93%.

By contrast with the 2 species and the suspected field hybrids these plants were markedly variable. This variability was analysed according to the scheme outlined above with the garden and glasshouse plants receiving separate treatment.