THE MULTIPLE TESTIS OF URODELES. 57 



ferred to regions of the testis emptying in the same autumn, in 

 some animals the regeneration of the lobules has been more de- 

 layed than in others. It is doubtful, too, whether a caudal lobe, 

 when it develops, occupies exactly the same extent of the germ-cell 

 cord as was occupied by the lobe which preceded it. Caudal lobes 

 of animals killed in September or October may show only a half 

 dozen lobules of mature spermatozoa, or they may be, on the other 

 hand, almost or quite equal in size to the lobes anterior to them. 

 The balance of metabolic processes in the animal doubtless deter- 

 mines the extent of movement of the spermatogenetic wave in any 

 season ; hence its cephalic progress in a caudally developing lobe 

 may be checked and the size of the lobe limited quite independently 

 of the exact time that had elapsed since the region was first emp-- 

 tied. The caudal lobe of Fig. 5, therefore, is not necessarily the 

 exact equivalent, in point of territory occupied, of the testis of 

 Fig. i, but may occupy a greater or less extent of the germ-cell 

 cord. 



In each year of sexual activity, as has been stated, the most 

 anterior lobe of the testis is shifted forward by exactly the extent 

 of the region maturing and extruding spermatozoa. It is inevi- 

 table that eventually this anterior lobe will reach the cephalic end 

 of the anterior germ-cell cord. By the time this occurs, however, 

 there will be one or more functional lobes established posterior to 

 it. Hence the animal's sexual activity is in no way interfered with 

 by the final disappearance of an anterior lobe. 



Such a disappearance or " running out " may be as easily demon- 

 strated as is the caudal addition of new lobes. In Fig. 15 is shown 

 an anterior lobe cephalad of which there is no cord of primary or 

 secondary spermatogonia. This condition marks the first stage in 

 the disappearance of the lobe. In Fig. 16 is represented a lobe 

 somewhat similar to the above, save that it is much shorter ; evi- 

 dently, in the preceding year, only a comparatively small number 

 of lobules of secondary spermatogonia had remained ahead of the 

 boundary plane when it was established, and had been held over 

 for later development. In Fig. 17 a more advanced stage in the 

 running out of a lobe is represented. Here the most anterior 

 lobules of the germ-cell cord have matured and extruded sper- 

 matozoa, and are found undergoing degenerative changes, sur- 



