XII.] CONJUGATION AND SEXUAL REPRODUCTION. 147 



perhaps a really different species from those of the variety 

 umvalens. In any case bivalens would be the ancestral form. 



While studying the last of O. Hertwig's works, the thought 

 occurred to me whether the fresh combination of ids in 

 Ascaris umvalens might not be brought about in a manner 

 different from that of the simple rearrangement of idants, and 

 I will take this opportunity of expressing the idea, in order that 

 its accuracy may be tested by the facts. The material for such 

 proof or disproof is not at present accessible to me ; for the 

 variety univalens does not seem to occur in south-west Ger- 

 many. 



In the sperm-mother-cells of Ascaris m. univalens four long 

 thin threads are formed from the chromatin distributed in the 

 nuclear network of the resting-stage ; these threads are arranged 

 so that they cross each other at one point and are there joined 

 together by means of a connecting cement-substance (' Linin '). 

 Thus they form an Ophiurid-like figure in which the body of 

 the Echinoderm represents the place where crossing occurs, 

 while the paired arms represent the halves of threads. 

 According to O. Hertwig, each of the threads then gradually 

 shortens itself until at length it resembles a short thick rod. 

 The four rods arrange themselves in two pairs, all four bases 

 being closely apposed, the spindle of the first ' reducing division' 

 is formed, and finally each daughter-nucleus receives one of 

 the pairs. 



Naturally, O. Hertwig was unable to follow these processes 

 directly, but he inferred them by combining the very numerous 

 stages observed. I should be inclined to look for a somewhat 

 different interpretation of the figures given by him, and would 

 ask whether the four threads which take the form of an 

 Ophiurid, are converted into the rods, not merely by shorten- 

 ing, but by the simultaneous /ws/ow of two half threads just as 

 if the paired arms of the Ophiurid, which lie side by side, were 

 to grow together. Many details support this view. First, the 

 connecting cement-substance at the point where the threads 

 cross certainly possesses some significance. If, however, the 

 nuclear rods arise by the shortening of the long threads only, 

 it would appear to have no meaning. Only if we consider that 

 it arises from the coupling together of different halves of threads, 

 would it possess a meaning, as will be immediately seen. If the 



L 2 



