1892.] NATURAL SCIENCES OF PHILADELPHIA. 373 



of organs, and the impressment of one set into the service of 

 another to such a degree that the morphologist will find little in an 

 original type to divide into sections. Monotropa w?i('/?ora represents 

 a section where the calyx is free from the ovary. The ovarium 

 is described as being 10-grooved, but it is evident that this grooving 

 is the result of ten staminal scales which have become adherent with 

 the ovarium. To describe the flower properly we might begin with 

 the 5-carpellate ovary. Though the '" stigma " is usually described 

 as being single and funnel-form, it would seem that a strictly cor- 

 rect form of expression, from the morphological standpoint, would 

 be that there were five stigmas, united by a thin membrane into a 

 circular, web-like disk. In the flowers before me there are ten 

 stamens, but these are certainly in two series of five each, one set 

 rather larger than the other. The upper and larger series alternate 

 with five of the scales, and press the anthers close up under the fine 

 stigmatic portions on the edge of the disk, and in this way effectu- 

 ally secure self-fertilization. The next series of five are shorter, 

 and always keep the anthers free from connection with the stigmas. 

 Both series of stamens, however, have the hairy filaments curved in 

 toward the ovarium as if they also would have become adherent 

 scales if they had had a fair chance. At the base of the ten sta- 

 mens, and alternate with them, are ten horn-like processes, evidently 

 each alternate one being slightly smaller than the other, represent- 

 inw two series which we may term either abortive stamens or abor- 

 tive petals, as suits best the morphological view. These, however, 

 curve outward and downwai'd instead of inward, and should pos- 

 siblv be classed with the corolliferous rather than the staminal 

 system which, as above noted, possesses an incurved tendency. From 

 the apex of these glands, however, a large quantity of sweet liquid 

 exudes and they might be termed " nectaries," if there were any 

 separate place for such organs in the morphological type. There 

 are five petals and it is singular that three of these seem always 

 to be twisted, folding over each other from right to left while 

 one petal is usually backed by the two adjoining (imbricate) 

 plates, the edges of these almost meeting behind the petal which 

 they enclose. Outside of these we have the remains of five sepals, 

 varying so much in size that occasionally but two can be recognized. 

 It will be seen that this conception, as fortified by observa- 

 tion, makes the flower far less of a departure from the usual types 

 of Ericaceje, and places the genus more in harmony with its fel- 



