THE CHLORANTHAL DIVISION. 



21 



Crassulacece. 



It appears not improbable that the broad scales opposite the lobes of the corolla in Boraginece are barren stamens, the filaments of which, if they were fertile, would be broader than those of 

 the fertile which are alternate with the lobes of the corolla ; — this really does occur in Echeveria among Crassulacece, where the stamens, which are alternate with the petals, have the usual narrow 

 filaments, and those opposite them have their filaments in the lower half as broad as the petals to which they are adherent nearly as high as the lower half of the petals * This dilated filament 

 has a lateral process on each side near the middle, and may therefore be regarded as the prototype of the scale, sometimes bifid, of the corolla in Boraginece. I regard Tetradiclis as near Crassulacece, 

 because it is an annual like some species of the British genera, is described as succulent, has a somewhat gyrate inflorescence, and a seed very closely resembling that of Penthorum, which has also 

 a flattened ovary and projecting placentae. Are not the long arms of the placenta a tendency to the same kind of displacement which Payer has shown to take place in Mesembryanthemum ? 

 See also Callitrichaceae. Tillcea, a part of the species of which are aquatic, forms a very close approach to Elatine except in its apocarpous ovary, but in other genera the ovary is partially syncarpous. 



Callitrichaceje & Batide^e. 



The species of Callitriche are small herbaceous plants, which are perhaps terrestrial rather than aquatic, as they grow more luxuriantly and produce much more seed out of the water on the 

 margins of rivers and ponds, than when they grow immersed, and they will grow with their usual vigour and fruit abundantly in a flower-pot. In habit they resemble Portulacece and Caryophyllacece, 

 C. verna when growing out of water, being much like Montia font ana, and C. pedunculata when growing on banks becomes tufted, much like a Sagina or Stellaria, and in the venation of their leaves 

 they also agree with Caryophyllacece (fig. 6). 



They further agree with Caryophyllacece, and with the nearly allied Tetragoniacece in the surface of their stems (especially when the plant grows out of the water), having a crystalline 

 appearance, which is owing to its being covered with crystalline glands (fig. 8). These have been very accurately described by Dr. E. Lankester (Linn. Proa, Vol. IL, p. 94), and require no further 

 notice, and are in all probability as he suggests, analogous to hairs. In Tetragoniacece, as in Callitriche, they occur thickly on the young branches, and more sparingly or scarcely at all on the 

 leaves, and I have no doubt about their identity, those of Tetragoniacece differing in being globular, so that they have not the stellate appearance of those of Callitriche. 



In C. verna there is a decided tendency in the flowers to become polygamous, as the stamen and ovary in the hermaphrodite flowers are neither of them smaller than usual. The stamen in 

 the hermaphrodite flower is remarkable for being posterior without any inclination to either side, a very rare occurrence ; and it is precisely hypogynous, the filament being attached partly to the 

 receptacle, and in a slight degree to the base of the short stalk which supports the ovary (figs. 6 & 7). It might be supposed that these hermaphrodite flowers were only apparently so and in 

 reality consisted of male and female flowers produced in the same axil ; but if this were the case there would in every probability be in some instances more than two bracts, but I have never seen 

 any trace of a third bract, and there appears no difference in the bracts as seen in the figure. When axils are 2-flowered they stand side by side, never as far as I have seen one directly before the 

 other and in every instance precisely so without variation. In Hippuris the stamen is anterior. The anther has been described as having but 1 cell, but in C. pedunculata it is 2-celled nearly up 

 to the period of maturity, so that the pollen can be removed from one of the cells and the firm intervening membrane distinctly seen (fig. 9). In its dehiscence, the anther of C. verna closely 

 resembles that of Campylostachys abbreviata presenting a semilunar 1 -celled appearance, as if the cells had become confluent, and this is equally remarkable in one species of Stilbe. 



The ovary is, as described by Mr. Babington, dicarpous, the dorsal suture of each carpel projecting inwards and uniting with the placenta, so as to make it 2-celled. The ovule is amphitropal, 

 the raphe not being more than half its length, and what is remarkable it scarcely increases in length as the seed ripens, so that in the matured seed it is scarcely more than a fourth of the length of 

 the seed itself; and another peculiarity is, that the nucleus incompletely distends the tunic of the ovule on the inner side next the placenta, and in the ripened seed a membranous portion of the 

 tunic is left in that situation, like a portion of a wing (figs. 10, 11, 12 & 13). The radicle is distant from the hilum by nearly half the length of the seed and not pointed directly to the 

 micropyle (fig. 14), so that together with the shortness of the raphe, the seed differs considerably from that of the Euphorbiacece with which Callitriche has been compared. The embryo is 

 nearly as long as the albumen and almost half its diameter, and the cotyledons are about one-third of the length of the embryo itself (fig. 14), so that in these characters it does not differ much 

 from Elatinacece and Crassulacece, Tetradiclis having a small quantity of albumen. And, another circumstance in which Callitriche agrees with Elatinacece is, that the achenium or half carpel 



* Described from a species in flower at Kew Gardens. 



