34 F. G. HEATHOOTB. 



stink-glands were formed comparatively late in the develop- 

 ment as invaginations of the dorsal plate, and I came to the 

 conclusion that they were not very deep-seated characters. If 

 this is so it is not difficult to understand that they may never have 

 heen developed in Polyxenus,but that the spines may have been a 

 sufficient protection. It is worth noting that the Archipolipoda 

 had spines, and not stink-glands. From all these points in the 

 anatomy of Polyxenus I am inclined to regard it, not as a 

 recently formed link between the Chilopods and Chilognaths, but 

 as an animal which has preserved certain traces in its anatomy 

 of its descent from a common ancestor of the two classes, 

 such ancestor being related to the Archipolipoda. I consider 

 it as confirming my view that the Myriapoda are descended 

 from a Peripatus-like form, and as opposing their descent from 

 Thysanura. I am fully aware that insomuch as I investigated 

 Polyxenus with a definite idea, I have probably taken a one- 

 sided view of the points I worked at; but I hope that my 

 work will induce others to investigate this animal more fully, 

 and thus increase our knowledge of the various questions 

 suggested by Myriapod anatomy and development. 



My work was entirely carried on in the Cambridge Morpho- 

 logical Laboratory. 



Literature. 



1. J. Bode. — "Polyxenus lagurus," 'Zeitschr. fiir die gesammten 



Naturwissenschaften,' 1877. 



2. Pabke. — "Anatomiedes organes reproduoteurea des Mjriapodes," 'Ann. 



des Sci. Nat.,' 4 ser., 1855. 



3. H. Gbenachee. — "Ueber die Augen einiger Myriapoden," 'Arcliivfiir 



mik. Anat.,' 18. 



4. Latzel. — ' Die Myriapoden der oesterreichisch-ungarischen Monarchie,' 



Wien, 1880. 



5. F. G. Hbathcote. — "The Poat-embryonic Development of Julus ter- 



restris,' Phil. Trans. Roy. Soc.,' 1888. 



