12 VASCULAR CRYPTOGAMS 



tube ; or probably the more correct homology is that of the nucleus of 

 the antherozoid with the * generative ' nucleus of the pollen-grain. The 

 most important difference between Cryptogams and Phanerogams lies 

 in the mode in which this contact between the male and female elements 

 is brought about. In Flowering Plants it takes place by the penetration 

 into the embryo-sac, through the micropyle of the ovule, of the extension 

 of the inner coat of the pollen-grain known as the pollen-tube, excited 

 into activity by the viscid secretion of the stigma. The pollen-grain 

 has therefore in the first place ' to be conveyed from the anther to the 

 stigma ; and the various parts of the flower of Flowering Plants are all more 

 or less concerned, directly or indirectly, with arrangements for facilitating 

 the conveyance of pollen. In Flowerless Plants, on the contrary, contact 

 between the male and female elements is effected by the protoplasmic 

 contents escaping from the male cells and coming directly into contact 

 with the oosphere in consequence of an independent power of motion 

 imparted to these naked masses of protoplasm (antherozoids) by the 

 vibratile cilia with which they are provided. This impregnation always 

 takes place in water or moisture, and no external agency is needed to 

 bring it about. Hence the absence from all Flowerless Plants of any 

 conspicuous compound organ analogous to the flower of Phanerogams. 

 The true homology of the pollen-grain of Phanerogams appears to be 

 with the microspore of the heterosporous Vascular Cryptogams, notwith- 

 standing the fact that the contents of the microspore break up into a 

 number of antherozoids, each capable of impregnating an oosphere : 

 while the pollen-grain, as a rule, emits only a single pollen-tube. Here, 

 as in other phenomena, we are guided to the true homology by compar- 

 ing the highest Cryptogams with the lowest Phanerogams, the Gymno- 

 sperms, which form a connecting link between them and Angiosperms. 

 In all Gymnosperms, as in some Angiosperms, the pollen-grain is 

 divided into several cells, only one of 'which (very much larger than all 

 the rest) emits a short pollen-tube. The large fertile cell in the pollen- 

 grain of Gymnosperms corresponds to the larger fertile portion of the 

 microspore of Selaginellaceae, to the entire contents of the microspore 

 of Marsileaceae, to the terminal cell of the germinating filament in 

 Salviniaceae, and to the antherid of the isosporous Vascular Cryptogams. 

 In Angiosperms the sterile cells of the pollen-grain of Gymnosperms 

 appear to be sometimes entirely suppressed, and the pollen-grain 

 becomes unicellular. The contents of the pollen-grain of Angiosperms, 

 together with the intine or inner coating of the grain, are therefore 

 homologous with the antherid of Cryptogams. Seeing that the motile 

 antherozoids have to be conveyed to the oosphere through the medium 

 of water, it is convenient for both antherid and archegone to be freely 



