440 MEBISTIC VARIATION. [part i. 



seen one with 8. Individuals with 4 arms occur, but are much 

 less common than those with 6. I have seen Asterina gibbosa 

 with 4 rays, and a specimen (Scilly) given me by Mr S. F. Harmer 

 has 6 rays, of which 2 are a little nearer together than the others 

 (suggesting division of a ray). Mr E. W. MacBride tells me that 

 he has seen several 6-rayed specimens of this species. Mr E. 

 J. Bles kindly tells me that he dredged a 4-rayed Porania 

 pulvillus in the Clyde estuary. There appeared to be no trace 

 of a fifth ray and the specimen was as nearly as possible sym- 

 metrical. 



The following cases exhibit special points. 



671. Asterias glacialis : specimen with 8 rays possessed 3 madre- 

 porites. Couch, J., Charlesworttis Mag. of N. R., 1840, iv. p. 34. 



(372. Asterias rubens : G-rayed specimens frequent atWimereux. In 

 several of these there are two sand-canals terminating at a common 

 madreporite. Giard, A., Comptes rendus, 1877, p. 973; cp. id. C. R. 

 soc. bid., 1888, p. 275. 



673. Partial division of an arm is fairly common in Asteroids, but less 

 common I believe than the total variation in number, though I know 

 no statistics on this point. For a figure of Asterias (Hippasterias) 

 equestris L. with a bifid arm, presenting no appearance as of regener- 

 ation see Tiedemann, Zeitschr.f Phys., 1831, iv. p. 123, Plate 1. 

 The two following are peculiar cases. 



6 / 4. Cribrella oculata : one of the arms bearing a branch, not as a 

 radius, but about (in dried specimen) at right angles to the normal arm, 

 the property of Prof. C. Stewart, who kindly shewed it to me. 



675. Porania pulvillus, Gray (a Starfish): Specimen 5 cm. in diame- 

 ter, having five short rays. The ray opposite the madreporite when 

 viewed from the aboral surface is seen to be distinctly bifurcated at 





1/ 



v^i.'i\\ & 





\ Mm* &*& a *~-s 





s. 







Fig. 134. Porania pulvillus, No. 675, having the arm opposite the madreporite 

 abnormally divided as shewn at x and y. (From a sketch kindly sent by Prof. 

 Hbbdmak.) 



