Saxifraga. | SAXIFRAGEA. 101 
large series of forms, some of them being near to S. Geum 
(sensu lato), others merging into S. wmbrosa—vide Pl. III., 
figs. 9-23, & Pl. VI, figs. 1-6. 
Many botanists have long regarded 8. hirsuta as a hybrid 
form. As far back as 1831, Hooker wrote in his Brit. Flora, 
2nd Ed., p. 192, ‘‘ my friend the Rev. W. T. Bree who has 
cultivated and studied the Saxifrages very assiduously, says 
that it [S. hirsuta] is certainly a hybrid between the preceding 
and the following [S. Geum and 8. umbrosa].”” This opinion 
is supported both by the known distribution of S. hirsuta 
and by the fact already referred to under S. elegans, that 
so many of its peculiarities are intermediate in character. 
To definitely decide whether this plant was really a hybrid 
of S. Geum and 8S. umbrosa, experiments along two distinct 
lines were undertaken by Prof. H. Dixon, F.R.S. of Trin. 
Coll. Dublin, at the suggestion of the author of the present 
Flora, and continued for many years to their completion in 
1914. Firstly, an endeavour was made to produce by the 
crossing of S. Geum with S. umbrosa, the intermediates which 
occur so frequently in a state of nature ; secondly, it was 
sought by the analysis of a natural intermediate (S. hirsuta) 
by self-pollination, to produce forms approximating on the 
one hand to 8. Geum, and on the other to S. umbrosa. Since 
each plant requires two years to reach maturity from 
germination, it is evident that no conclusive result could be 
obtained for several years. 
1. ARTIFICIAL PRODUCTION OF INTERMEDIATES. 
Representatives of both S. Geum and S. umbrosa were 
gathered in natural situations in Kerry. Photographic leaf 
outlines of the S. Gewm parent selected are given on Pl. IV.,* 
figs. 1, 2. As will be seen, they fully comply with the 
Linnean definition of this plant “ foliis reniformibus den- 
tatis.” Leaf outlines of the S. wmbrosa parent are shown 
on the same plate, figs. 3, 4. 
These plants were taken into cultivation at Trin. Coll. 
Botanic Garden, Dublin, and with due precautions were self- 
pollinated in order to ascertain their purity as types. The 
seeds thus produced were collected and sown, and 45 
* Prate IV. Hybridising Experiments, Synthesis of Intermediates, Geum 
parent—figs. 1-2 ; umbrosa parent, figs. 3-4 ; leaves from resulting plants, 
figs 5-16 ; figs 17-18 are derived from fig, 6, self-fertilised ; figs. 19-21 are 
derived from fig, 7, self-fertilised, 
