igi2. Notes. 205 



NOTES. 



BOTANY. 



A Note on the Robertsonian Saxifrages. 



I have never been able to follow the splitting of Saxifraga umbrosa and 

 5. Gcum into the several species or varieties which at various times have 

 been described — S. serratifolia Mackay, S. elegaus Mackay, S. hirsnta L., S. 

 punctata Haw. These are all forms connected by many intermediates. As 

 has often been pointed out they are characteristic of places where S. 

 umbrosa and 5. Geum grow together. I cannot regard them as connecting 

 links between the two, which seem to me to be quite distinct species ; and 

 I think that the botanist who studies them in the field will soon come to 

 tne conclusion that they are of hybrid origin. The " splitter " who wants to 

 coin half-a-dozen " species" quite as good as those named above, need only 

 go to certain places in Kerry, such as Connor Hill, to find ample material 

 for his ingenuity ; and it seems to me that the describing of such a " new 

 hybrid " as 5. Geum var. serrata x umbrosa var. serratifolia, as my friend 

 Mr. Marshall has recently done [Joiirn. Bot., June, 1912) is a mere waste 

 of time. Several " new hybrids " have from time to time appeared spon- 

 taneously in my garden, where the two species — wild plants from Galway 

 and Kerry — grow together ; and I am now growing a beautiful series of 

 intermediates, representing every gradation from t^-pe umbrosa to type 

 Geum, collected by the roadside at Connor Hill Pass, a few months ago. 

 In the present note I merely wish to emphasize certain circumstantial 

 evidence which appears strongly in favour of the hybrid origin of all the 

 intermediate forms ; of course the only conclusive proof of the same 

 will be their production in the garden by artificial crossing, and I do not 

 doubt that this can be done by everyone who takes the trouble. 



In the first place, S. umbrosa is, in nature, an extremely stable species 

 In Galway and Mayo, for instance, one may examine thousands of speci- 

 mens without detecting any noticeable variation, except that occasionally 

 the toothing of the leaves is accentuated. In Donegal and on the Blaskets, 

 likewise, where 5. Geum is absent, the plant remains quite typical. But 

 here is an interesting point. In all Galway and Mayo, just one small 

 patch of S. Geum is known — namely, on the great cliff on Clare Island, 

 where 5. um.brosa is also abundant ; and within a few yards of it, and 

 nowhere else, there occur a few plants of a form just intermediate between 

 the two species. Similarly, when we cross the Blasket Sound towards 

 Brandon, the advent of 5. Geum is accompanied by the appearance of a 

 swarm of puzzling " varieties." It seems to me that the only chance of 

 escape from the hybrid theory lies in the assumption that S. Geum is a 

 very variable species. Unfortunately, we have no place in Ireland where 

 5. Geum is present and 5. umbrosa absent, in which this theory can be 

 tested : but the occurrence of every gradation from Geum to typical 

 umbrosa to my mind puts this theory out of court. Of course each par- 

 ticular hybrid keeps " thoroughly constant " in cultivation (a point put 

 forward by Mr. Marshall in favour of the subspcciftc rank of S. serratifolia) ; 



