102 PILULARIA. [BOOK i. 



ring detaches itself from the stalk at one end, straightens, and 

 carries up with it the spikes of fructification, whose connec- 

 tion with the stalk is then destroyed. The spikes are at first 

 enveloped in a mucous membrane, and are composed of two 

 sorts of bodies closely packed together, and considered by 

 M. Dunal to be ovules and anthers. These bodies are some- 

 times intermixed, sometimes stationed separately from each 

 other. The so called ovules are little white semitransparent 

 bodies, surrounded by a sort of projecting hood, beyond which 

 a narrow papilla projects : this papilla is always turned 

 towards the anthers. The latter are little flat parallelopi- 

 pedons, rounded at the two ends ; they consist of a mem- 

 branous sac of great tenuity, in which are found numerous 

 grains of spherical or elliptical pollen. (Ann. Sc., n. s. vii. 227. 

 t. 12, 13.) M. Fabre is represented as having proved experi- 

 mentally that the latter impregnate the former ; and he has 

 traced the ovules from their first impregnation to their com- 

 pletion, and seen and described their germination. (Id. ix. 

 115, t. 13.) It appears that no trace of embryo is discoverable 

 in the ripe seed. 



In Pilularia the organs of reproduction lie in hairy oval 

 cases, or sporangia, whose interior is divided into four cells 

 filled with bags arranged in four lines as in parietal placentae, 

 some containing a germinating body or sporule, others filled 

 with a powdery matter. The first have been regarded as 

 pistils, the latter as anthers. But Mr. Valentine has shown, 

 in a very detailed memoir, that the so called anthers are 

 merely abortive spores, as I long ago suggested. (See Lin- 

 naean Transactions, vol. xviii. tt. 34 and 35.) 



Salvinia and Azolla have been the subject of some elaborate 

 observations by Mr. Griffith, (Calcutta Journal, vol. v.) He 

 regards them as having true sexes, the male being certain 

 necklace-shaped threads found at an early stage, in contact 

 with what he denominates an orthotropous ovule. But strange 

 to say, this so called ovule, instead of giving birth to an em- 

 bryo, becomes the parent of reproductive bodies of two totally 

 different kinds, having not even the smallest resemblance the 

 one to the other, although the matrix out of which they are 

 evolved is identical at an early period of the organisation. 



