HYDROMEDUSAE 



30s 



quadrangular, with only an indication of four perradial lobes. In the collection from 

 1 90 1 (20 November) there are three very interesting young specimens showing the 

 first development of the canals. In two of them there are eight simple radial canals 

 without any branches (Fig. 2) ; two canals run from each of the four corners of the 

 stomach to the margin of the umbrella, but these two canals are unequal. One of them 

 proceeds in a perradial direction to the umbrella margin, and at its termination there is a 

 large tentacle ; the other canal from the same corner of the stomach runs in an oblique 

 direction to a point of the margin in the middle of the space between two perradial 

 tentacles ; at its termination there is a small rudimentary tentacle bulb, and above this 

 there is an exumbrellar nematocyst cluster, whereas such are not found above the 

 four perradial tentacles. In later stages the eight "main" radial canals are all alike in 

 appearance, but from these young stages it is evident that only four of them are truly 



Fig. 2. Willia mutabilis. The radial canal 

 system of the earliest stage, showing four 

 primary, perradial canals and four canals run- 

 ning to the four interradial tentacular bulbs. 



Fig. 3. Willia mutabilis. The radial canal 

 system of an early stage, showing commence- 

 ment of torsion of the umbrella ; also a single 

 branch to each of the eight radial canals. 



primary radial canals, whereas the four others are lateral branches of the former. 

 Presumably therefore a still younger stage exists, with only four radial canals, as in 

 young specimens of Proboscidactyla. 



In the third specimen of this sample a torsion of the umbrella has just commenced 

 (Fig. 3) ; the course of the four primary canals is not exactly perradial in relation to the 

 stomach, and the corresponding tentacles are a little displaced, in a clockwise direction 

 when the medusa is seen from above. The four interradial basal bulbs have developed 

 into tentacles which, however, are distinctly smaller than the four perradial ones; 

 moreover, there is a lateral branch to each of the eight radial canals, issuing somewhat 

 nearer to the umbrella-margin than to the stomach, all to the same side ; at the termina- 

 tion of each of these lateral branches is a very small basal bulb, and above each of these 

 a cluster of nematocysts. 



There are several slightly older specimens in which the two radial canals issuing 



