470 



Variation in " Ophiocoma Nigra " 



20 



20 



Tliis liiie is shuwn in Diagnuii L, and is si-en to represent very c4osely the 

 observed facts. 



30 



DiAORAM I. Regression of Arm Length on Breiidth of Disc. 

 40 50 _ 60 



70 



30 



40 50 



Ann Length in MilUmetres. 



60 



70 



This rcsult, that the arm-length is fivc times the width of the disc, is almost 

 a mein between JeffVey Bell's corresponding fignre 7^ and Forbes' 3 to 3^. 

 Thcir calculations were in all probability madc froin a few speeimens, while mine 

 are from one thousand. It nuist, however, be remenibered that niy nieasurements 

 are froin si^ecimcns which had been lying in alcohol for soine weeks ; but even 

 although there inay have beon some contiactiou, it is ncjt likely that the ratio of 

 the breadth of disc to the length of ray would thereby be sonsibly affected. 



As it appears from Table III. that the entire ränge of Variation of the disc- 

 breadth is from 4 to 17 mm., it might readily be supposed that the size of the 

 mesh of the net used in dredging had much to do with the fact that no speeimens 

 were found whose diameter measured less than 4 mm. Not much importance, 

 however, can be attached to this objection, because neai-ly all the animals collected 

 were brought to the surüvce clinging to blades of oarweod. 



From the totals in Table III. we may form a polygen of frequencies. Diagi-am II. 

 gives the curve whirh best fits this. The constants of this curve are : 



niean 10-3.")7 mm. 

 mode 10628 mm. 



0-= 2i20(; iinn. 



/x-, = 4-5222 

 /i, = - 2-2366 

 M4= GO-2962 



80 



15 



10 



80 



