BALFOUR—NEW SPECIES OF RHODODENDRON. 57 
It must be borne in mind that this discussion is tentative. 
We have as yet seen too little of the plants and have too little 
material to work with for a thoroughly critical analysis of the 
problem. But it seems to me worth while to give this exposition 
of facts because of their bearing upon the question of Rhododen- 
dron-hybridsin nature. Had we no knowledge of Rh. nertiflorum 
or of Rh. dimitrum, the forms Rh. dimidiatum and Rh. diphro- 
calyx would, by all the canons of systematic botany, be recog- 
nised as two species (possibly to some as microforms of one), 
without near affinity with either Rh. callimorphum or Rh. 
habrotrichum. Jf Rh. callimorphum and Rh. habrotrichum had 
been undiscovered, there would have been no hesitation on the 
part of systematists over describing Rh. dimidiatum and Rh. 
diphrocalyx as two species (or microforms of one) allied to but 
quite distinct from Rh. neriiflorum and from Rh. dimitrum. 
With all the forms before us, the suggestion of hybridisation as 
the method of origin of Rh. dimidiatum and Rh. diphrocalyx 
is legitimate. If this is the true history, these forms must be 
hybrids that have arisen in nature; and this has support in 
the first appearance of Rh. dimidiatum amongst seedling plants 
of Rh. callimorphum, of Rh. diphrocalyx amongst seedling plants 
of Rh. habrotrichum. Rh. neriiflorum was introduced to culti- 
vation in 1910, Rh. habrotrichum in 1912, Rh. callimorphum in 
1914. I do not know if Rh. dimitrum is yet in cultivation, and 
the time that has elapsed since any of them came into our 
gardens—perhaps I should put it since the first record of their 
flowering, the all-important Rh. dimitrum being, as I believe, 
still absent—is too short for the production of flowering plants 
in progeny from artificial crossing. We have therefore here a 
critical case in which synthesis may confirm or otherwise the 
conjectures of analysis which the method is inadequate to 
establish on the material available. If Rh. dimitrum is not 
yet in cultivation, it is a plant to be resolutely sought for in 
view of its importance as an element in the question that has 
to be answered. Meanwhile cultivators of Rhododendrons who 
interest themselves in hybridisation, and who have in their 
collections plants of the species required, should endeavour, by 
making such a series of crosses as neriiflorum x callimorphum, 
neriiflorum x habrotrichum, callimorphum x nertiflorum, habro- 
trichum x nertiflorum, to find out for us what are the characters 
of the first-crosses that come from such parentage ; and should 
also, by crossing directly and reciprocally Rh. dimidiatum and 
Rh. diphrocalyx with their reputed parents, try to discover 
through Mendelian segregation evidence of the parental relation- 
ships of the two forms. Speculation is frequent upon the part 
that hybridisation in nature has played in the production of 
