ORIGEN OF METAMERIC SEGMENTATION. 79 



into meso- and meta-nephros; the raeso-nephros becomes 

 connected with the male generative organs, and loses its ex- 

 cretory function, while the metanephros persists as the func- 

 tional kidney. I have, however, fully discussed the evolution 

 of the Vertebrate excretory system in my papers already quoted 

 on their development, and need not refer further to it here, 

 except to point out that there is every reason to believe that 

 the nephridia were originally segmental, one for each somite, 

 that this segmental arrangement is, with the specialisation of 

 the kidney, soon lost as it is in other organs. 



On the Structures known as Primitive Streaks. 



I may conclude this paper by a short review of these 

 structures. 



(1) They are always connected with the formation of the 

 mesoblast. 



(2) They are never, so far as I know, found in free larvae. 

 They are confined to the embryonic phase of development, and 

 are only found in animals which undergo a considerable part 

 of their development in the egg ; in other words, only in eggs 

 well-stocked with food yolk, or in eggs which have lost the 

 food yolk. On the other hand, a primitive streak is not 

 universally presentin such cases, e.g. Cephalopoda, Elasmo- 

 branchii, Amphibia, Crustacea. 



(3) They are always median and unpaired in their origin, 

 but may in later development become grooved and present 

 traces of a bilateral structure. 



(4) They are always caused by rapid proliferation of cells 

 apparently from the epiblast. 



(5) Their position seems to vary in different animals. 



In Vertebrata, when present, the primitive streak is placed 

 mainly behind the blastopore (according to StrahP not entirely 

 so in Lacerta, but this is not quite clear from his figures). 



In Peripatus it is placed behind the blastopore, and, when the 

 blastopore has divided, behind the hinder division (fig. 4). 



1 ' Arch. f. Anat. u. Pbys.,' 1S82. 



