Sawxifraga.] SAXIFRAGEA. 99 
Mountain, Connor Hill, Brandon Mountain, &c., and which 
are unknown in those parts of Ireland where S. wmbrosa 
grows alone, as on the mountains of Waterford, Tipperary, 
Galway, &c. This fact by itself would raise doubts as to 
its claim to specific rank, but when it is found moreover 
that S. elegans as well as many other of these various forms 
combine in different degrees the respective characters of 
S. Gewm and S. wmbrosa,* a strong presumption of their 
hybrid origin is established, independently of the results 
obtained by Prof. Dixon in the experiments which are re- 
ferred to under 8. hirsuta on p. 101, &c. 
Of all the Robertsonian forms, however, the one which 
has given rise to the most discussion is the Sazifraga hirsuta 
of Linné. In separating this oval-leaved plant from his 
S. Geum, Linné himself appears to have felt misgivings, as 
he states in his Species Plantarum, that it is only too close 
(“‘ affinis nimium”’) to S. Geum. His selection of the name, 
too, was rather unfortunate, as the degree of hairiness is 
quite a subordinate and variable character in all these forms. 
Smith, in Hng. Bot., Vol. II., p. 263, speaking of S. hirsuta 
* As showing the influence of S. umbrosa, the following characters of 
S. elegans may be noted—the absence of a cordate base to its leaves, the 
eT flattening of the petiole, and the presence of a few red spots on its 
potals. 
Although the flowers of the Robertsonian Soxifrages in Kerry examined 
by the writer do not vary very perceptibly witk the different leaf forms, 
the number of the spots on the petals follows fairly general rules which 
may be summarised as follows—In reniform-leaved S. Geum there are rarely 
more than 1 or 2 coloured spots to be found on each petal in addition to 
the irregular yellow blotches common to the lower half of the petals of all 
the Robertsonian Saxifrages in Kerry. In forms where the leaves tend to 
become orbicular or ovate, and the cordate base to disappear, a few red 
or orange spots are usually present ; this is seen in forms like S. elegans 
and S. hirsuta, &c. The more the base of the leaf tends to merge into the 
petiole, as in several of the forms connecting S. hirsuta with 8, umbrosa, 
the more do these spots appear to increase in number, reaching their maxi- 
mum in typical S, wmbrosa, where they occasionally amount to 11 or 12 in 
addition to the yellow basal spots, There appear, however, to be many 
exceptions to this colour scheme ; in about 10 p.c, for example, of the 
S, umbrosa plants examined, usually the most spotted member of this 
Robertsonian group, the red or orange spots wore only one or two on each 
petal, and in a few instances were completely absent. ae! 
It appears to be impossible to discover any definite distinction between 
the two species either in the form or arrangement of their petals or in the 
shape of their capsules. Nevertheless this gencral statement appeers to be 
justified, that the petals of S. wmbrosa are usually broader than those of 
S. Geum and its allied forms, and that hairy leaves and those netted with 
purple underneath are very rarely scen in S, wmbrosa or its forms though 
commonly present in S. Gewm. ‘The capsules of all the various forms 
examined were filled with apparently fertile seeds, 
