OF THE MttLLERIAN DUCT OF AMPHIBIANS. 515 



lies, the ccelom is very much reduced by the existence of secondary tissue, which 

 surrounds the alimentary canal and envelops the lungs, covering the entire body- 

 wall except a small ventral region. 



Opposite the apparently closed funnels of the anterior nephrostomes there are slight 

 diverticula of the ccelom containing the glomeruli ; but there are no diverticula dorsal 

 to the lungs. These appear a few sections further back. 



On both sides, the anterior nephrostomes, though apparently closed, open backwards 

 into the body-cavity by means of narrow superficial tubes, whose apertures are consider- 

 ably more posterior than the original funnel.* 



Between the new mouth of the first nephrostome and the second nephrostome the usual 

 band of thickened epithelium runs. It covers the entire dorsal face of the diverticulum, 

 but is particularly distinct laterally, i.e., where it meets the attachment of the gut. 

 Where the alimentarv canal becomes free from the lateral walls the thickening is found 

 to run far round ventrally. It can be traced into the middle diverticulum as far as the 

 peritoneum has not been encroached on by the secondary attachments, so that it appears 

 that the heightened epithelium of the pronephros region has simply been divided 

 anteriorly by the backward growth of the attachment of the alimentary canal. A 

 wedge-shaped region of the epithelial plate has been covered over, and two tapering 

 anterior processes connected by a posterior transverse band persist. The attachment of 

 the gut ceases immediately posterior to the second nephrostome, and the two bands form 

 one broad plate, which, however, soon runs down to a narrow band that passes along 

 the segmental duct towards the posterior end. 



My next stage (" E ") of Salamandra atra is found in a specimen 37 mm. in length. 

 The pronephros has in it undergone extensive degeneration, and there is only an almost 

 median remnant of the posterior part of it. The dorsal diverticula seem practically 

 closed up by the attachment of the gut, and there is no sign of nephrostomes or of the 

 heightened epithelium between them. Beyond the secondary attachment, however, i.e., 

 in the middle diverticulum, a narrow band of thickened epithelium runs back. It passes 

 obliquely round the posterior of the attachment of the gut — behind the remnant of the 

 pronephros — and then back along the segmental duct as a more or less distinct ridge of 

 cells. The segmental duct is in many places quite indistinct. 



To illustrate a stage intermediate between " D " and " E " I have figured sections of 

 a Salamandra maculosa of 29 mm., showing the cutting off of the dorsal band of 

 thickened epithelium and the connection of the unsuppressed remnant with the narrower 

 and more ventral band that here, as in Salamandra atra, runs along the middle 

 diverticulum, and serves as the foundation of the anterior end of the Miillerian duct. 

 In fig. 12 the much reduced band opposite the second nephrostome is shown to be 

 separated from the deep and narrow lateral band by the attachment of the gut. Fig. 1 3 

 shows the third section further back, and illustrates how the two bands form one where 

 the secondary attachment ends. 



* For further particulars as to the closing of the nephrostomes, vide p. 519. 



