14 
gersen has misunderstood the structure of the gonophore. He 
describes the two tentacles as “bent up and concealed inside the 
umbrella”, and he States: “The umbrella contains four distinet, 
wide and simple radial canals, connected distally by a ring-vessel • 
a velum is indicated ...” (p. 23). I have sectioned two well-devel- 
oped gonophores (see textfigs. 7 and 8), and I have found that the 
Figs. 7 and 8. — Ichthijocodium sarcotretis. Transverse (fig. 7) and longitudinal (fig.'8) 
scctions of gonophores. — ex. exumbrella ; sub. subumbrella ; man. manubrium ; t. 
tentacular bulb ; v. velum ; r. end. radial strings of endoderm. 
radial “canals” are four solid strings of endoderm going from^the 
base of the manubrium to the beil margin, terminating in triang- 
ular tentacular bulbs, two of which are mueh larger than the 
others. The endoderm strings are fairly thick, broadly elliptic in 
cross-section. The transverse section (textfig. 7) shows that they 
project more inwards (to the subumbrella side) than outwards, 
forming four longitudinal ridges on the subumbrella. There are no 
tentacles. What J u n g e r s e n considered to be tentacles, “bent up 
and concealed inside the umbrella”, are really, I am sure, the 
thick longitudinal ridges, mentioned above. The velum is very 
broad, turned outwards between the tentacular bulbs. In the inter- 
radial spaces between the endoderm strings the two layers of ecto- 
derm are only separated by a thin mesosarc; there is no endoderm 
lamella. The lack of tentacles in the two gonophores, sectioned by 
me, and the other characters by which they differ from Junger¬ 
sen s description, cannot be explained by the specimens belonging 
