Mr. J. Miers on Ephedra. 249 



that in Gnetum the male and female flowers are found in distinct 

 whorls on the same node, I had suggested the possibility that 

 in Ephedra both sexes might prove to be developed in the same 

 common spikelet, in which the male flowers in the lower whorls 

 had fallen away before the female flowers became developed in 

 the terminal whorl — a supposition rendered more probable by 

 the fact that male and fructiferous spikelets are sometimes found 

 on the same plant. But the changes shown in the gradual 

 development of the ovary and fruit of Welwitschia render the 

 above supposition improbable; and by analogy we may now 

 form a tolerable conjecture of the nature of the female flower in 

 Ephedra. From these data we may infer that the two ovaria 

 developed in the terminal pair of involucels are deficient of a 

 corolla — a circumstance which sometimes occurs in Euphor^ 

 biacecE, where the male flowers are provided with both calyx and 

 corolla, while the ovary is destitute of any floral envelope. 



The application of the term " cone " to the flowering heads of 

 Welwitschia and Ephedra is calculated to mislead many persons 

 in regard to the affinity of the Gnetacea; for they bear little 

 analogy to the cones of the Coniferce. They are more properly 

 spikelets, because they bear regular petaloid sessile flowers along 

 a common axis, much after the manner of a spike of Plantago ; 

 and they offer more claims to this category than the spikelets 

 of Myrica, the aments of Betula, or the spicated inflorescence of 

 many other genera. 



The structure of the male flowers and the mode of inflores- 

 cence in Welwitschia present a striking resemblance to those in 

 Ephedra, both showing an advanced state of floral development. 

 Dr. Hooker considers the ovule in the female flower to be deficient 

 of any carpellary covering, and therefore gymnospermous ; but 

 the circumstances he has demonstrated tend rather to evince 

 that it is enveloped in a distinct carpel. The important fact of 

 the existence of hermaphrodite or polygamous flowers in this 

 family serves to throw much light on this point. It is shown in 

 pi. 6. fig. 14 that Welwitschia (besides its floral envelopes) pre- 

 sents a monadelphous ring of regularly formed stamens sur- 

 rounding an ovary constituted in the usual manner of angio- 

 spermous plants — that is to say, with a simple style and stigma 

 surmounting an oblong 1 -celled carpel containing a single erect 

 ovule, thus exhibiting a floral development and pointing to a 

 position in the system far higher than the gymnospermous 

 orders of Conifera and Ci/cadacece. But the ovule of the herma- 

 phrodite flower is always sterile, and it is only in such flowers 

 as are deficient of corolla and stamens that embryo-sacs are 

 formed in the ovule which admit of its fertilization ; and here it 

 is seen that the style becomes so far depressed that the stigma 



Ann. 6f Mag. N, Hist. Ser. 3. Vol. xi. 17 



