Saxifraga.] SAXIFRAGE. 97 
Coll. Dublin, of which a short account is given under 8S. 
horsuta on p. 101, &c. On Plates I. II. III. & VI. photo- 
graphic reproductions of a series of these natural leaf- 
forms are given, each leaf having been taken from a plant 
growing in Kerry, with one or two exceptions which are duly 
indicated. 
Unfortunately the material preserved in Linné’s own 
Herbarium is both scanty and unsatisfactory, and gives no 
help in determining the form or forms included in his 
S. Geum.* The description, however, of this plant in his 
Species Plantarum, 2nd Ed. (1762), p. 575, “‘ foliis reniformi- 
bus dentatis,” limits the type to plants with kidney-shaped 
leaves. Of S. Gewm so limited, hereinafter referred to as 
S. Geum (sensu stricto), two forms are met with in Kerry, 
one with rather large leaves and blunt, sometimes almost 
flat, crenations rather than serrations ; the other with leaves 
usually smaller, and with teeth more or less acute. In 
both forms the petioles are slender and nearly cylindrical, 
and like the leaves are more or less hairy. Some leaf outlines 
of the crenate form are given on PI. I.,t figs. 1-6; it is most 
probably the Robertsonia crenata of Haworth ; while figs. 
11-18 give leaf outlines of the serrate form, probably 
Robertsonia dentata of the same author. While both these 
crenate and serrate forms fully comply with the Linnean 
definition of S. Geum, and in their extreme states are easily 
separable from each other, plants of S. Geum (sensu stricto) 
may be found with teeth quite intermediate in character; 
a few examples of these are given on Pl. I., figs. 7-10. 
These reniform types, which are quite local in the county 
occur in the following localities—I. Uragh Wood by Lough 
Inchiquin, Cloonee valley, 1908 ; sparingly by the roadside 
in the upper Slaheny valley, 1904.—II. Sparingly on 
Beeown Mountain north of Sneem, 1903 ; rather sparingly 
in damp gullies near Coomakesta on the Waterville-Darry- 
* Mr. B. Daydon Jackson has kindly sent to the writer an exact outline 
tracing of a leat from a specimen labelled S. Gewm in Linné’s own Herbariun, 
preserved in the rooms of the Linnean Soc., Burlington House. The leaf 
sketched is longer than broad, not cordate, and would be better referred 
to S. hirsuta (sensu lato). Another leaf-tracing, from the same source, of a 
specimen of Linné’s S, hirsuta shows an ovate leaf tapering into a broad 
footstalk and appears to be nearly related to S. wmbrosa var. punctate. 
{+ Puate I. (} life-size)—S. Geum Linn. (Sensu stricto, i.e, Reniform).— 
Crenate forms, figs. 1-6. 
Intermediate forms, figs. 7-10. ! 
Serrate forms, figs. 11-18. 
