6 Mr. A. Alcock o?i the 



enough lias been said to enable us to understand the meaning 

 of these appearances. 



In transverse sections of a trophonema taken from a long- 

 gravid uterus in which the foetus, having used up all the 

 jolk, is now demanding other nourishment, there is, as has 

 been shown in the paper already quoted, little to be seen but 

 two opposed rows of bulb-shaped milk-secreting glands with 

 funnel-shaped mouths, separated by a vascular space. These 

 glands take the place of the more or less solid nests of cells 

 seen in the above-described sections of trophonemata from a 

 gravescent uterus. And in comparing this less mature with 

 that more mature stage we come to the conclusion that in 

 Pteroiilatoia micrura^ as Professor Wood-Mason and myself 

 have already shown to be the case in Trygon walga, the 

 secreting glands of the nursing-filaments, like the alveoli of 

 the milk-glands in Mammals, begin as solid nests of epithe- 

 lium, which, with the onset of active secretion, gradually 

 become hollow chambers by the breaiiing down and exfolia- 

 tion of their core. 



5. Consideratio7}s as to the Descent of the 

 Pteroplatcean Alliance. 



I hope before long to give a more complete account of the 

 embryonic history of Pteroplatoia micrura, from which perhaps 

 it may be more permissible than it can be from the meagre 

 facts just recorded to attempt to retrace the pedigree of the 

 Trygons. But on account of the recent revival of interest in 

 the phylogeny of the Batoidei it will, I trust, be considered 

 pardonable to touch a few points, from all of which we can, 

 without straining, bring these embryos into the field of vision. 



It is impossible to see these little embryos without in the 

 first place being struck by their shark-like form ; and when 

 next attention is fixed upon the gill-openings — their con- 

 spicuous dorsad extension and their relation to the prolonged 

 pectoral fins — one is immediately reminded of RMna. Indeed 

 all that is needed is to straighten out and flatten the pectoral 

 fins and to depress the head in order to get a strong resem- 

 blance to that interesting intermediate form. Or, if we leave 

 the pectoral prolongations untwisted and imagine them in 

 this condition fused with the head, we get a remarkable like- 

 ness to Ceratojjiera and Dicerohatis. 



The descent of the Trygonidas from a shark-like ancestor 

 is of course, from what is well known of Roja and Torpedo^ 

 only what would be expected ; but 1 do not know whether or 

 not the suggestion that the line of descent passes (1) through 



