10 FOSSIL MEDUS. 
however, to compare the lobed margin with that of figs. 7, 7a of Pl. XVII. 
The four oral arms of Brooksella are not known to have been subdivided 
(see Pl. II, figs. 8, 8a); nor is there any trace in the fossils of a system of 
sucking pores, corresponding to the sucking frill on the axial side of the 
arms of Archirhiza. That such sucking pores existed is highly probable. 
The transverse section of the recent Cannorhiza connexa Haeckel is of 
value in interpreting the structure of Brooksella. Thesection shows the main 
radial canals extending from the central stomach to the margin, pillar canals, 
Fic. 2.—Archirhiza primordialis Haeckel. View from below; twice the natural size. The two paired arms of the 
left half are cut away. In the center of the arm disk the regular cross of the mouth seam is visible, the four branches of 
Maes fork at the ends of the brachial trunks. op, perradial sensory button; oi, interradial sensory button; cc, ring 
arm pillars, buccal stomach, oral arms, brachial canals, and sucking mouths. 
In the fossil forms (PI. II, fig. 8b) the oral arms, brachial canals, base and pil- 
lar canals, radial canals, and central stomach are more or less clearly shown. 
These are all general features, but taken in connection with the.character of 
the umbrella, they serve to classify the Brooksellidze under the Discomeduse, 
but not to establish their family relation with any of the living families. 
The living forms of medusze that resemble Brooksella and Laotira most 
clearly in their habits appear to be Polyclonia and Cassiopea (pp. 6, 7). 
