3i6 MOSSES AND FERNS chap. 



mother cells of the spermatozoids, and has much more dense 

 contents than the outer cells, but all the chloroplasts remain in 

 the latter. A third wall now forms in the upper peripheral 

 cell, much like the first one in form, and cuts off a cap cell at 

 the top. The young antheridium at this stage consists of four 

 cells — a central dome-shaped one surrounded by three others, 

 the two lower ring-shaped, and the terminal one discoid. These 

 outer cells are nearly colourless and contain -very little granular 

 contents, except the small chloroplasts, which are mainly con- 

 fined to the surface of the inner walls. 



The divisions in the central cell are at first very regular. 

 The first one is always exactly vertical, and is followed by a 

 transverse wall in either cell which strikes it at right angles, 

 and next a third set of walls at right angles to both of these, 

 so that whether seen in cross-section or longitudinal section, 

 the central cells are arranged quadrant-wise. Successive bi- 

 partitions follow in all the cells until the number may be a 

 hundred or more, but the number is usually much less, about 

 thirty-two being the commonest. The regular arrangement of 

 the sperm cells soon becomes lost, and they form a mass of 

 polyhedral cells with dense granular cytoplasm, and large nuclei. 

 A nucleolus is visible until the last division, after which it can 

 no longer be distinguished ; otherwise the nuclei show no pe- 

 culiarities. The transformation of the nucleus into the body of 

 the spermatozoid proceeds here as in other Ferns that have been 

 examined, but I was unable to satisfy myself that so large a part 

 of the forward end of the spermatozoid is of cytoplasmic origin, 

 as Strasburger ((ii), IV, p. 115) asserts. The fully- 

 developed spermatozoid describes about three complete coils 

 within the globular sperm cell, and does not lie coiled in a 

 single plane, as in the Hepaticae, but in a tapering spiral (Fig. 

 174, D). The very numerous long cilia are attached at a 

 point a short distance back from the apex, and as Buchtien 

 ((i), p. 38) showed, cover a limited zone, although hardly 

 so restricted as he figures. 



From the investigations of Shaw (2) and Belajeff (5, 6, 7), 

 it is evident that the cilia arise from a blepharoplast. Belajeff 

 considers the blepharoplast in the Pteridophytes, as well as in 

 the Bryophytes, to be a centrosome ; but Shaw believes that the 

 blepharoplast is an organ sni generis, and of quite different 

 nature from the centrosome. 



