of the 0/-^H/M- r^'Piliilaria gloljulifera. 485 



projection, or a general swelling of that part (Tab. XXXIV. fig. 12.). By dis- 

 section, however, we can observe this process at an earlier period ; and I find, 

 upon removing the conical projection, that the first evidence of germination 

 having commenced is an appearance of cellnlarity within tiie unsupported or 

 valved poi'tion of the internal membrane, which is transparent ; and I now 

 find, for the first time, a very delicate pellucid membrane lining- the whole 

 cavity of the sporule, and having the cells, which give the appearance just 

 mentioned, either lying on its external surface, or forming that portion of it 

 which lies beneath the valves. From the appearances, and the impossibility 

 of separating the cells from the membrane, I am inclined to believe that the 

 cavity formed by this membrane is completed by the cells, or, in other words, 

 that the sac is partly membranous and partly cellular (Tab. XXXI\". fig. 10 

 & 11.). It is possible that this last-described membrane may exist before ger- 

 mination begins, notwithstanding the numerous unsuccessful dissections which 

 I have made to discover it, the failure being owing to its extreme delicacy ; 

 but I am pretty well satisfied that it is a product of germination, because I 

 have not the slightest difficulty in demonstrating it after that has commenced, 

 nor is there the slightest trace of it in any stage of the development of the 

 sporule. However this may be, it is quite certain that fresh cells are gradu- 

 ally formed on the external surface of the cellular part of the sac, and that 

 the valves of the third membrane arc very soon ruptured and graduallv 

 turned back by the growth and i)rotrusion of this button-like cellular germ 

 (Tab. XXXIV. fig. 9 & 13.). The enlarging cellular mass then distends the 

 conical projection, unfolding the plicae of that body, and at length appears 

 externally with four of its cells projecting beyond the general mass, and com- 

 pressed into a cpiadrangular form by the pressure, as I imagine, of the apex of 

 the cone, the aperture in which is quadrangular. These projecting cells soon 

 harden, and acquire a reddish-brown hue, and in the advanced stage of ger- 

 mination appear like a brown quadrangular space, which I should have little 

 hesitation in referring to the above cause, did I not find several similar spaces 

 on the germinating spornles of Isoëtes lacustris, which I could not refer to such 

 an origin : it must be observed, however, that I have not seen the earlier stages 

 of germination in Isoi-tes. Soon after the exposure of the entire germ, which 

 is effected by the reflection of the valves and conical membrane over the side 



3 s 2 



