472 ' MR. E, T. BROWNE ON * [Mar. 17, 
belongs to the Gymnoblastic group of Hydroids. The Medusoid 
Willia stellata has been always considered a Leptomedusa and 
was placed by Haeckel among the Cannotide. AsI did not know of 
any clearly proved case of a gymnoblastic hydroid having a medusa 
which must be classified with the Leptomeduse, I examined by 
means of sections the exact positions of the gonads. It is one of 
the characteristic features of the Anthomeduse to have the 
reproductive cells on the wall of the stomach, and of the Lepto- 
medusze to have the reproductive cells upon the radial canals. 
When I first saw Willia I certainly regarded it as a Lepto- 
medusa. At the base of the stomach six broad canals are given 
off ; these run along a short peduncle and over the top of the 
sub-umbrella; along this portion the gonads are situated; at the 
point where the gonads terminate the canals suddenly decrease in 
diameter and become somewhat inconspicuous. The question arose 
whether to consider the broad part of the canal upon which the 
gonads are situated as a prolongation of the stomach or as a part 
of the radial canal itself. To determine this point, I cut several 
series of sections of meduse belonging to the second and third 
stages. All the sections clearly show that the endoderm-cells of the 
stomach are continued, without any change in size or shape, along 
the whole length of that portion of the canal upon which the gonads 
are situated. At the point where reproductive cells terminate the 
canal suddenly decreases in diameter, and the endoderm-cells 
become very small and flat. I think that portion of the canal 
upon which the gonads are situated may reasonably be regarded as 
a prolongation or lobe of the stomach, and that the true radial 
canal commences at the end of this lobe. 
The sections also show reproductive cells at the base of the 
stomach itself, before it branches into the six lobes. 
I think this medusa is as interesting as the hydroid is remark- 
able. It shows a kind of transitional stage between the Antho- 
meduse and Leptomeduse. The continuation of the endoderm- 
cells, without change, into the lobes, and also the continuation of 
the reproductive cells from the base of the stomach itself along the 
lobes, strongly point to Willia belonging to Anthomedusz, and not 
to the Leptomeduse. 
DIsTRIBUTION :— 
Hydroid Form. 
England—Ifracombe, Hincks *. 
Medusoid Form. 
Scotland—St. Andrews, M‘Intosh. Oban, Forbes. 
England—Plymouth, Garstang and E. T. B. Fowey, Peach. 
Falmouth, Cocks. Penzance, Forbes. Ilfracombe, Gosse. 
Treland—Valencia Island, #. 7. B. 
