VARIATION IN A JELLYFISH, 1053 



three quadvalits. This specimen had interra.dial, adradial and 

 subradial tentacles in two of these, whilst the third quadrant 

 carried only a single adi'adial one *. 



Giinther has shown that in Limnocnida [9J, tentacles of a 

 particular order are often fully formed in one quadrant before 

 there is any trace of them in the others. Browne also has called 

 attention to the fact that in Podocoryne carnea [7] the young 

 medusfe do not always leave the hydroid colony with the same 

 niunber of tentacles, all have foiu" perradial ones but the number 

 of interradial tentacles varies, some having two or three instead 

 of four, and one specimen he noticed to have a single one only. 



I do not think that in the case of the asymmetrical medusse of 

 Mmrisia to which I have just referred, we are dealing with any 

 retardation in the development of the tentacles in certain 

 quadi-ants ; some of these individuals were of considerable size 

 and the gonadial divei^ticula well developed [cf. text-fig. 226), yet 

 certain quadrants showed no signs whatsoever of secondary 

 tentacles, although other quadi-ants possessed tentacles of the 

 third or even fourth order. 



These five asymmetrical medusfe clearly indicate that each 

 quadrant is capable of forming secondary tentacles qviite 

 independently of the other quadrants of the bell. That this can 

 occur in a radially symmetrical animal is distinctly stated by 

 Bateson [15] who remarks, that in radial series " as in Linear 

 Series, Variation, whether Meristic or Substantive, may take 

 place either in single segments (quadrants, sixths, etc.), or 

 simultaneously in all the segments of the body." This statement 

 was based on observations made on two medusae, Clavatella 

 {Eleutheria) proUfera and AitreUa aurita. 



Clavatella is a medusa which normally bears a single ocellus 

 at the base of each of its six tentacles. Claparede [16] has called 

 attention to the fact that these ocelli are sometimes doubled ; 

 this duplicity may occur at the base of a single tentacle or 

 occasionally at the base of each tentacle instead of one. 



Bateson also quotes the observations made by Romanes [4, 5] 

 on Aurelia aurita. In this form, in addition to changes symmetri- 

 cally carried out in the whole disc, one or more quadrants may 

 vary independently. Thus one specimen is figured in which two 

 quadrants are normal (i.e., each possesses one generative organ 

 and a set of radial canals) but the other half-disc is divided into 

 three. Similarly a particular quadrant may possess two sets of 

 organs or even three, the other three quadrants being normal or 

 nearly so. 



In addition to the forms descinbed above, three other medusae 

 showed abnormalities of a quite different type, which, although 

 known to occur in other genera, I consider worth recording. 



In the first of these specimens, which was normal as regards 



* This specimen is one of those not included in the Table published in iny 

 original paper on Mcerisia Jyonsi. 



