140 Mr. W. T. Thiselton Dyer. 



Fig. 5. 



Spreading. 



I was much, struck to find amongst a magnificent series of speci- 

 mens, kindly sent me by Messrs. Sutton, forms with the segments 

 of the corolla spreading instead of reflexed (tig. 6). I have received 

 even more striking examples from Messrs. Hugh Low. This is 

 remarkable because, as I have already pointed out, the latter is a 

 distinctive generic character in Cyclamen. Although the alteration 

 in the appearance of the flower is enormous, the structural change is 

 slight. It is merely a matter of direction of growth. It amounts, 

 however, to the loss of a generic character and a reversion to a more 

 generalised type. The change is therefore essentially atavistic. 



I was unable to obtain from Mr. Martin any explanation of how 

 this particular variation had come about, but he informed me that it 

 had been of frequent occurrence. Spreading flowers had always been 

 destroyed as departing from a desirable type. More recently, on 

 account of their orchid-like habit, they had taken the popular fancy, 

 and had b3en preserved. 



