372 PROCEEDINGS OF THE ACADEMY OF [1892. 



cially Mr. John H. Redfield, brought me generously large amounts 

 of fresh jjlauts, their daily collections. One of these, which I 

 had no subsequent opjiortunity to see growing, was Dalibarda 

 repens. My observations are therefore, wholly from these few plants, 

 though the facts have been subsequently confirmed by Mr. 

 Redfield.^ 



The single flower on the four to five inch scape, was found wholly 

 staminate, the stamens being very numerous. The scape branched 

 at the base, having what we might term a pedicel of an inch or so 

 in length, recurving and bearing small cleistogene buds. These 

 were found to contain but five short stamens, and the anthers, pol- 

 liniferous, pressed down on the stigmas. These bud-bearing jjedi- 

 cels grew into the rotten leaves beneath the foliage, and matured 

 the seed vessels beneath the surface, just as many violets do, 

 bringing to mind that Michaux''s name of D. violceoides, was still 

 more expressive than even that author supposed. 



As many of these subterranean seed vessels were ripe, it shows 

 that cleistogamy commences quite early in the season, how early 

 has yet to be ascertained. The eai'liest ones, it Avould seem, must 

 have the flower on the 5-inch scape perfect, as the descriptions given 

 by authors, no doubt, were drawn from these, and are given as with all 

 the floral organs perfect. My rough notes, made on the spot, say, 

 " the female or fertile flowers, with becoming modesty, withdraw into 

 privacy beneath the ground, while the foppish male flowers seem to 

 have no office in life to perform but to make a show of themselves." 

 It may be remarked here, that in many cases of plants bearing 

 cleistogene and fertile flowers they at some time produce what are 

 regarded as normally perfect flowers, and these are supposed to be 

 for the purpose of enabling the j)lant to get a chance to bear cross- 

 fertilized seed. But so many of these are wholly infertile, while 

 this Dalibarda is absolutely masculine, that it would seem that the 

 fact may be capable of some other explanation. 



On Some Morphological Distinctions in the Genera of 



Ericace.e. 



The sub-orders of Ericacete divide naturally into the free or the 

 adherent calyx with the ovary, but an examination of various mem- 

 bers of the order indicates a suppression or multiplication of series 



^ As these pages are going to press, Mr. Fedfield notes that Dr. Gray has 

 recorded the finding of Dalibarda repens cleistogene by Mr. Pringle. 



