While 110 (lefinile conclusions can be drawn from the few larvtv of 1 lo- 

 ioliinrians hilherto known, and likewise the Ciinoids are onl of (|ueslion 

 in I his connection haviiii* no line |)elai>ic larvie. the Ophiuroiii hirvie iiave 

 an imi)orlanl bearini* on the i)rol)leni. Larva' l)eloni<ini> to the genera 

 Ophiolhri.v, Ophiopholis, Ophiavlis ami Amphiiini are known bnt only one 

 of each genus. They do not alTord so very characteristic dilTerences as 

 one should expect, since they belong to three dilTerent families, and 

 those of Ophiopholis and Ophiavlis are more dilTerent than should l)e 

 expected after the apparently near relation of these two genera. Then 

 within the genus Ophiunt we know with certainty the larv?e of the two 

 species Oph. alhida and Icrliinild. and tliese are so dilTerent that one would 

 rather think they nuisl belong to dilTerent families. The larvte which are 

 with a considerable degree of probability referred to Oph. affinis and 

 Homahphiuia (/Hida (Koehler) again are very dilTerent from each other 

 as well as from the two named above. These facts are thus decidedly 

 opi)ose(l to the conclusions to be drawn from our previous knowledge of 

 the Echinoid and Asteroid larva". Considering, how^ever, that onr knowledge 

 of the natnial affinities of the vast nund)er of Ophiurids is still rather 

 unsatisfactory, even in s|)ite ol the more iccent attempt at a natural 

 classification of this group by Ma I sumo to, it seems to me that the 

 facts known of the Kchinoid and Asteroid larvcT outweigh those of the 

 Ophinroid larva>. and lead to the conclusion that these larvae, on the con- 

 trary, tend to i)rove that there is something wrong with our classification 

 of the Ophiuroids. 



In spite of these OphiuraAmviv it seemed then a legitimate conclusion that 

 the Echinodei-m larva' are really of considerable classilicatory value, the 

 larvtv of nearly related forms agreeing in their main characters; the opposite 

 conclusion would then be equally legitimate, that when the larvae of two 

 forms, appaiently nearly related, prove to be essentially dilTerent, those 

 forms are not in reality nearly related. The study of the larval forms will 

 then alTord a very important test for the value of our classification of 

 the adult forms. 



It is evident thai the knowledge of a very much larger number of 

 Kchinoderm larva' than the comparatively few known till now is reijuired 

 for placing the idea of the classilicatory value of these larva' on a more 

 firm ba.se. Nearly all the researches hilherto iikkK' on the development 

 and the larval I'oinis of f'.chinoderms were carried onl on species occurring 

 in the iMiropean ami .North American Seas, only a few forms from the 

 tropical seas of .Vmerica having more recently been made the object of 

 study (fen 11 en I). Hut, moreover, it is mainly a few selected types 

 which have, over and over again, been studied, so that even of Ihe com- 



* 



