114 
distinguish it. A root from Snowdon was not altered in the 
garden at the time of flowering, in the following year; but as 
it was soon afterwards destroyed I had not ‘opporthnigg to 
make further observations. In the autumn following, the 
leaves were found very large, having from five to nine spread- 
ing segments, or doubly 3-parted. In the flowering state 
they corresponded with the variety 2. On the other hand, 
wild specimens from Twll-di, with terminal flower-stalks, 
rather belong to the variety y. 
In some of the wild specimens from Wales, I observe the 
anthers to be minutely pointed; in other cases their lobes are 
spreading below. In a few instances the filaments are 
scarcely, if at all, dilated at the base. The stigmas appear to 
be of uniform structure, accidentally variable in size and 
direction; equally so on the same individual plant as in the 
different varieties. In one or two of the Snowdon specimens, 
I find the ribs of the leaves, especially of the foot-stalks, very 
strong and prominent. The calyz-segments are, in every 
instance, 3-ribbed. The flowers in a dried state are either 
quite white, or, in some varieties, yellowish or cream-coloured, 
indicating almost a specific difference, and the ribs on the 
petals are in some individuals much more conspicuous than 
usual. The seeds, so far as I have observed, are always 
rough, either ovate or roundish. Embryo nearly central, with 
2 cotyledons. 
39. Saxifraga affinis—The character and description would 
hardly induce me to admit it as a species—if cultivation should 
be found not materially to affect it, I would not altogether 
refusé to receive it, but if the plant has been only partially 
examined, and the description made from one or two speci- 
mens only, I should have serious doubts concerning this plant. 
40. Saxifraga incurvifolia.—If the obtuse incurved segments 
of the leaves should be found constant, I may consider it dis- 
tinct from S. hypnoides and the varieties, but it would perhaps 
not be very easy to distinguish it from S. cespitosa, with 
which the calyx-segments seem to correspond. The «upright 
shoots" may vary by culture or otherwise, as is the case with 
S. cespitosa, and, as it appears, also with S. elongella. 
