428 MOSSES AND FERNS chap. 



short, curved bodies. Their number could not with certainty 

 be determined. The nucleus passes through the various 

 karyokinetic phases, and the directive spheres occupy the poles 

 of the nucleus spindle when at a later period they divide so 

 that each daughter nucleus has two of them. After the 

 daughter nuclei have assumed the resting condition these can 

 no longer be distinguished, and what their fate is must for 

 the present remain undecided. The resting nuclei, as in other 

 cases, show no nucleolus. Fig. 222, F to J, shows the 

 earliest stage in the differentiation of the spermatozoid, and this 

 corresponds exactly with what I have observed in various 

 Ferns, and differs somewhat from Buchtien's figures of corre- 

 sponding stages. The nucleus, which is not noticeably lateral 

 in position, shows a narrow cleft upon one side. Seen in 

 profile (Fig. 222, F, i), one side projects somewhat more than 

 the other, and becomes the anterior end, which later becomes 

 thinner than the posterior part. I was unable to see that this 

 forward part behaved differently with regard to the nuclear stain 

 employed from the hinder part, nor could I satisfy myself of 

 the presence of the cytoplasmic anterior prominence which 

 Strasburger ^ figures in the Ferns. Staining with the mixture 

 of fuchsin and iodine-green, recommended by Strasburger, gave 

 indifferent results, both in the younger stages and the free 

 spermatozoids. In microtome sections, where the spermatozoids 

 were very strongly stained and the cytoplasm almost colourless, 

 the nuclear structure was unmistakable nearly to the extremity. 

 It is not impossible that the extreme forward end may be 

 cytoplasmic ; but if so, it forms but a very insignificant part of 

 the fully-developed spermatozoid. The cilia (Fig. 222, C) are 

 evident long before the spermatozoid has attained its full length, 

 and are shorter at first than later on. There seems little doubt 

 that they are direct outgrowths of the forward end, and lie close 

 to the convex surface of the body, so that they are easily 

 overlooked. 



The body rapidly elongates and becomes quite homo- 

 geneous, but this does not occur until a comparatively late 

 stage. The nucleus is here somewhat flattened to begin with, 

 and the coils of the spermatozoid lie nearly in the same plane 

 and resemble a good deal those of Marattia, except that they 

 are larger. The protoplasm enclosed within the coils is con- 

 1 Strasburger (11), vol. iv. PI. III. Figs. 26, 27. 



