THE CRASPEDOTE MEDUSA OLINDIAS AND SOME OF ITS NATURAL ALLIES. 7 



exumbrellar ectoderm lying about the tentacle roots is conspicuously thickened, 

 and contains black pigment granules and nettle-cells. 



The velar tentacles are very long, slender, and contractile when fully formed, 

 but are found in all stages of development from a mere bulb to a long contractile 

 filament armed with numerous stinging batteries. They are always found close to 

 the base of the velum, and their endoderm is directly continued into that of the cir- 

 cular canal, without the intervention of what may be called "roots." In younger 

 examples, of about 20 millimetres or a little more in diameter, the velar tentacles are 

 all in the condition of bulbs, exactly like the basal bulbs of Gonionema ; but in mature 

 specimens (70 millimetres or more in diameter) some of the velar tentacles are exceed- 

 ingly long and contractile, and are armed with numerous nettle batteries in the form 

 of incomplete rings. These incomplete rings are arranged all with the open point 

 turned towards the oral side, so that there is formed a longitudinal groove on the inner 

 side of the filiform velar tentacle. The ectodermal muscle fibres of these tentacles 

 are at least three times as thick as those of the exumbrellar tentacles and are cross- 

 striped. This is in entire accord with their great contractility in life. These filiform 

 velar tentacles are never very nmnerous, there being some ten or so in an example 

 of about 60 millimetres in diameter. In the aquarium these tentacles are dragged 

 along passively when the animal swims forwards, but when at rest they are loosely 

 laid out on the bottom and occasionally contracted. I believe they are the principal 

 organs for capturing prey. When strongly contracted, they tend more or less to 

 form a spiral. In an example of 15 millimetres in diameter there were about 100 

 velar tentacles, all in the condition of bulbs, and in one of 75 millimetres some 325, 

 of which a dozen or so were filiform. 



The velar tentacles are very different in structure in the bulbular and filiform 

 conditions. In the bulbular condition the ectoderm is so clogged with nettle-cells in 

 all stages of development that it is hard to recognize the individual ectoderai cells. 

 It is also to be remarked that none of the numerous nettle-cells are found fully devel- 

 oped. The endoderm is a direct continuation of the external part of the endodermal 

 wall of the circular canal, with which it presents the same histological characters, 

 and the mesogloea is very thin. In the filiform condition the mesogloea is tolerably 

 thick, the endoderm cells are large and vacuolated, though less so than those of the 

 exumbrellar tentacles. The stinging batteries contain numerous fully developed 

 nettle-cells, and between the ectodermal cells close to the mesogloea are found numer- 

 ous nettle-cells in the later stages of development. These are placed with their long 

 axes parallel or only slightly oblique to the length of the tentacle, and are, in my 

 opinion, on the way to their destinations, the earlier stages having been passed in the 



