PHANEKOGAMIA : FLOWERS. 357 



tinous membrane ; and that each four lie thus in a thicker gelatinous 

 parent-cell, which is manifestly such from its origin. But the so-called 

 special parent-cells are not easily distinguished either from the parent- 

 cell, or from one another, or from the enclosed pollen granules at this 

 period. Their origin may, indeed, be other than that assigned by Na- 

 geli. The following appears to be equally probable. In the parent-cell 

 are formed, according to the laws of development, four pollen-cells ; 

 whilst these expand, the granular contents of the parent cell are gradu- 

 ally dissolved into gelatine, in which the pollen-cells then lie imbedded. 

 By this pressure of the expanding pollen-cells, a portion of the gelatinous 

 substance is thickened or condensed into a membrane around and enclos- 

 ing each, and thus forms the so-named special parent-cells. 



Where, on the contrary, the parent-cells first divide into two cells, 

 two special parent-cells are really formed, but in the same manner as I 

 have just described of the pollen-cells ; and in each of these special parent- 

 cells two pollen-cells are formed in the same way. It would be in har- 

 mony with this that a distinct pollen binarium occurs, as in the Podo- 

 stemacece, which indicates the closer relation of the pollen-cells. Only 

 continued and careful investigation can decide whether Nageli's excellent 

 observations, as he has related them, are to be regarded as complete, or, 

 according to the hypothesis I have given, to be brought into agreement 

 with other processes of cell-formation. 



On the other matters mentioned in the paragraphs, I have only to add 

 that I have found Nageli's observations respecting the order in which the 

 parent-cells and the special parent-cells are dissolved, and upon the ex- 

 ternal pollen membrane, a secreted product of the pollen-cells, perfectly 

 correct. 



The perfect pollen granule consists, as has been said, of the essential 

 pollen-cell, which in plants flowering above the water is clothed with 

 a peculiar secreted layer. This always forms an uniform membrane 

 lying immediately upon the pollen-cell, not unfrequently in a double 

 layer, on which usually appear all kinds of strange projections (the 

 first product of the secretion). Most frequently these little ridge-like 

 projections are connected together in a reticulated form, and often 

 give the external membrane the deceptive appearance of being com- 

 posed of cells, which the history of development proves not to be the 

 fact. The spaces between the meshes of this net are frequently partly 

 filled with transparent jelly (in Iris and Passiflord). Sometimes 

 these articularly- connected ridges present very regular, definite 

 areas, giving, in the excessively varied form and arrangement of the 

 pollen grains, the most elegant and beautiful appearance : this is 

 especially the case in the Passiflorce. Sometimes, again, the pro- 

 jections appear as minute points, cones, papillae, curves, or like little 

 towers, either scattered upon the surface or very regularly arranged 

 there (e. g., most elegantly in many of the Composites, Scorzonera, 

 Tragopogon, &c.). The substance of this secreted layer is usually 

 yellowish, more rarely tinted with green, blue, or red, and by con- 

 centrated sulphuric acid it is only very slowly destroyed (requiring 

 one or two days), and during this the acid often gives it a 

 Burgundy-red colour. 



In all pollen granules the external pollen membrane exhibit? 



A A 3 



