HARGITT. 



[VOL. I. 



In Gonionemus, as in most Hydromedusae, the number is 

 four. But in the specimens studied in this connection it was 

 not rare to find three or five or six. This is the more remark- 

 able in that Bateson 1 has been able to record but rare instances 

 of such variation among Hydromedusae, instancing those noted 

 as among the most striking illustrations of the " discontinuity 

 of meristic variation." 



A fourth series consisted in sections of the medusa in a 

 horizontal plane, as indicated in Fig. 4. As will be perceived, 

 this is clearly the most crucial of the entire series, since the 

 animal is divided into such parts that if regeneration occur at 



B 



FIG. 4. 



FIG. 5. 



all it must be dc novo, from tissues devoid of any direct 

 identity of form or function. But crucial as is the test, it was 

 none the less successful. In Fig. 4, A and B represent the 

 central and marginal portions of the divided medusa. In Fig. 5, 

 A and B represent the corresponding portions after regeneration 

 had occurred. A in this figure needs no special explanation. 

 That which is first regenerated in this portion is the marginal 

 canal. Next following this is the regeneration of the tentacles. 

 As will be noticed from the figure, they are but rudimentary or 

 bud-like. The unprecedented hot weather which occurred just 

 at this time made extremely difficult the preservation of speci- 

 mens in a healthy condition under the artificial conditions of 



1 Bateson : Material for Study of Variation, pp. 424 et seq. 



