SPERM ATOZOIDS IN TORREA VITREA. 77 



and present themselves, as is usual, in different degrees of 

 development. At first they are perfectly diaphanous, smooth, 

 and manifestly homogeneous, without any trace of an enve- 

 loping membrane. The dimensions attained to by them in 

 this condition reach to as much as l-16th of a millimeter in 

 length, and 1-2 3rd of a millimeter in breadth. 



At this epocli they may be seen to exhibit two grooves, 

 crossing each other at a right angle, and whose direction 

 has not appeared to me to present any constant relation 

 with the form of the mass itself. It is probable that this first 

 form of division may in some sort be accidental, for I have 

 only very rarely noticed it. 



The number of grooves soon increases, and they become 

 more marked and deeper, and the mass, after having presented 

 a surface subdivided into large irregular lobes, assumes a mul- 

 berry-like aspect, and ultimately becomes completely granu- 

 lous. During the time that these phenomena are being 

 manifested, the mass continues to increase in volume, and in 

 its ultimate condition it is sometimes l-12th of a millimeter 

 long by nearly 1-1 6th of a millimeter broad. 



The masses when a little further advanced soon split up, 

 and the tail of the spermatozoids is then apparent. The 

 spermatozoids continue to adhere to each other for some time 

 longer by their bodies, as well as to the granulations not yet 

 transformed ; ultimately they are gradually separated. 



At the moment when the spermatozoids separate themselves 

 from the minute masses, of which they constitute a part, their 

 body is almost fusiform, and perhaps not more than 1- 100th 

 millim. long, and l-300th millim. thick. But they grow 

 during the time they remain in the midst of the fluid 

 which bathes them, the body and the tail elongate ; and besides 

 this the former increases considerably in its transverse dia- 

 meter. Among spermatozoids quite mature, some will have 

 attained to a length of l-60th millim., and breadth of 1- 1 50th 

 millim. 



I have long since remarked the analogy presented between 

 the progressive breaking up of the spermatogenous masses 

 and that of the vitellus. Numerous observers, it is well known, 

 have confirmed what I have written on this subject since 1845, 

 but it is a point upon which I have found myself continually 

 at discord with some who have been specially engaged in re- 

 searches of this nature. 



In Germany, more especially, almost every naturalist who 

 has spoken of the development of the spermatozoids has 

 applied, in this department of physiology, the cell-theory of 

 Schwan. The spermatogenous masses, in their eyes, have 



