of Eclinhurgh, Session 1883-84. 
763 
To trace the supposed transformatiou in the reverse order, it is 
thus. The original annelid-like ancestor first of all changed its 
position in locomotion, so that its nervous system became dorsal 
instead of ventral. Then, by the folding over of its median nervous 
plate to form a tube, the two series of neuropodia were brought 
together, coalesced, and formed the dorsal fin of the fish. 
The animal having obtained a new anus some distance from the 
end of the body, the posterior part of the intestine disappeared, 
being no longer funtional; and by the consequent reduction of this 
part of the body, the posterior part of the two series of notopodia 
coalesced in the median line and formed the anal fin. Anteriorly 
the successive notopodia united in two separate regions to form the 
pectoral and pelvic fins, and in the intermediate region disappeared. 
In the course of my observations on the development of floating 
Teleostean ova, carried on in the month of June last, I was im- 
pressed with the incompatability of Dohrn’s ingenious theory, with 
the existence of a pre-anal ventral median fin in the larvae of one 
species. The larvae in question were hatched in the Scottish 
Marine Station at Granton, from transparent floating eggs obtained 
by means of a fine tow-net in considerable numbers, about 50 miles 
east of the Isle of May. I do not know at present to what fish the 
ova belong, but as the ova and embryos are well characterised, they 
may be identified at some future time with eggs taken from some 
ripe adult fish. The yolk in both the developing ovum and the 
hatched larva is extremely pellucid, and divided into separate por- 
tions with polygonal outlines : there are no oil-globules in the yolk, 
the notochord is composed of a single column of vacuolated cells in 
the formed embryo, and in the larva the anus is situated at some 
distance from the yolk-sac and close to the posterior extremity of 
