THE ANDR@CIUM IN COCHLIOSTEMA. 207 
lopment of this remarkable flower. I have the more hesitation in 
doing so, as the course of development in the andrecium is so 
different from that described by Payer in Zradescantia or Com- 
melyna*. For the sake of comparison I add a brief abstract of 
the conclusions arrived at by the eminent French savant. Payer 
describes the sepals as developed in succession, the latest being 
the anterior one or that superposed to the “bractée mére;” the 
petals are described as simultaneous in their appearance—one 
placed internally, one externally, and one, half internally, half 
externally. So far my observations are quite in accordance with 
those of Payer; but in the case of the andraecium there are ma- 
terial discrepancies, as will be obvious from a comparison of what 
has been previously stated with Payer’s account of the develop- 
ment of the stamens in the regular-flowered Tradescantia. The 
androecium, according to this botanist, consists of six stamens in 
two rows: the stamens of each whorl appear all at the same time; 
but the members of the inner series, those superposed to the petals, 
appear first, before those placed in front of the sepals, a wholly 
exceptional occurrence. It is unnecessary to say more as to the 
development of the other parts of the flower, or as to the inflor- 
escence, as my observations on these points coincide with those of 
Payer. I may, however, indicate that although Cochliostema in 
the adult state is, by reason of its structure, far removed from 
other genera of the same family, yet there are points in its 
conformation which appear to connect it with other genera of the 
same family. Forinstance,in Commelyna itself, of the six stamens 
the three posterior only are fertile, 7. e. one of the outer row and 
two of the inner, exactly as in Cochliostema. In some species of 
Aneleima on the other hand, as also in Dichespermum, the three 
outer stamens, placed in front of the sepals, are fertile, the three 
inner ones being sterile. Dithyrocarpus has all six stamens fertile. 
The hairs on the filaments of Tradescantias and Commelynas may 
probably be looked on as having some relation to the way in which 
the flower gets fertilized; but whether they have the same office 
as the petal-like hoods of Cochliostema is a matter of doubt. 
Again, the peculiarly dilated connective of many Tradescantias is 
clearly analogous to the expanded filament of the plant under 
consideration. 
In conclusion, then, it would appear from what has been above 
* Payer, ' Organogénie,' p. 663, t. 140. 
