38 R- T. Günther 



depresseci as iu Cla rateila prolifera, the creeping larva oü Eleuthena. 

 An iini)ortaut fcatnre Is the freqiient })resence of an in-i)ushiug or 

 'Mimple" iu the centre of the exumbral surface, from which, in 

 typical cases, four centrifugai grooves proceed interradially to the 

 margiu. A siuiihir grooving of the exumbrella is known in other 

 Cladüuemidae, cspecially duriug contraction e. g. Gemvmria sagittaria 

 Hck. Very frequently one of the grooves is so exaggerated as to 

 uotcli the uml)relhi niargin deeply; but the marginal notch is hy no 

 uieaus so Constant in its presence as to justify us in regarding it, 

 with Claus, as diagnostic of the genus Mncstra. The marginal notch 

 is showu in figs. 1, 2, 36, 39, 40, but was absent from some in- 

 dividuals. I have never seen a specimen with the four meridional 

 grooves developed with such faultless radial symmetry as those 

 drawn by Müller & Gegenbauh and Lankester (figs. 37, 38). 

 All specimens examiued by me were asymmetrical in some respect, 

 and within certaiu limits exhibited a great variety of form and 

 structure, a fact which is in great measure to be attributed to their 

 sessile mode of life. It should however be stated that the symmetry 

 of the velum and circular canal is not much affected by the asym- 

 metrical deep grooves or notch upon the surface of the exumbrella. 

 The central in-pushing in the middle of the exumbrella (figs. 1, 

 12) is, I think, to be regarded as homologous with the umbral canal 

 (Stielcanal) which places the gastric cavity of the developing Cras- 

 pedote medusa in communication with the gastric cavity of the 

 parent from which it lias budded. If this interpretation be the true 

 one, the in-pushing would be regarded by Hartlaub as homologous 

 with the umbral cavity (rfcheitelhöhle) of Eleutheria. 



The armament of the exumbrella. 



The exumbrella is well armed withbatteries of cnido])lasts which 

 are disposed in a ring all round the margin of the umbrella and 

 along four perradial streaks which extend centripetally from the 

 tentacle bases to the in-pushing in the centre of the exumbrella, 

 and sometimes continue down into it. There is a considerable 

 amount of variability in the exact distribution, width and length of 

 these streaks. To these cnidoblast streaks I have applied the name 

 of cnidacts^. 



* From -/.vtÒTj a nettle and 'oc/xi; a ray. 



