1916] Weatherby, — Staminody of Petals in Amelanchier 40 



which, under the microscope, proved to be indistinguishable in ap- 

 pearance from pollen-grains. 



I have since examined the readily available herbarium material of 

 the small petaled plant — about a dozen sheets in all, from Blue Hill 

 in Milton, Mass., from Nantucket and from various localities in Con- 

 necticut. Every one shows the same condition to a greater or less 

 extent. I have not seen the actual specimens from which Mr. Bick- 

 nell described J. nantveketense, but the fact that he speaks of the 

 petals as "often involute" 1 indicates that it is not unlike the rest. 

 The degree to which modification of the petals has progressed varies 

 in different individuals and in different flowers on the same shrub - 

 even in the same flower. Some are oblanceolate to obovate and show 

 only an inrolling of parts of the margin; others possess more fully- 

 formed anther-saes and have a narrow claw and a short, sub-orbicular 

 blade, suggestive of filament and anther in outline, but petaloid in 

 texture; in still others, the lower part of the claw has the heavier 

 texture of true filaments; arid in a few, the whole has the texture of 

 the true stamens and can be distinguished from them only by the 

 broader and Hatter filament, the longer anther-sacs and the broad 

 connective, corresponding to the blade of the petal. 



Such a series of structures could, of course, be produced in the 

 comparatively common phenomenon of "doubling" the trans- 

 formation of stamens into petals. But in that ease one would expect 

 to find the usual number of normal petals in addition to the partially 

 developed ones, and a reduction in the number of stamens correspond- 

 ing to the number of the latter. Xo such condition appears. Whether 

 modified or not, the petals are five and no more; and the number oi 

 stamens, though variable, as usually in Amelanchier, averages about 

 the same as in normal flowers. It seems plain, then, that we have here 

 a case of transformation of petals into stamens, and that to this 

 teratological tendency is due the dwarfing of the petals in this puzzling 

 plant. What the underlying cause of the tendency may be, I cannot 

 now suggest. This note is published in tin 1 hope that others may be 

 moved to make observations of the plant in the field, to see if any 

 evidence of disease or other external cause can be found. 1 noticed 

 none in my one meeting with the growing plant; but 1 was not then 

 looking for it. 



East Hartford, Conn. 



i 1. r. 454. 



