ECHINOIDEA. II. n 



ttofhuridfe and Echinonietridæ^ or better still, to read the chapter oii the classification of the Diadeni- 

 atids in m\- paper on the Siani-Echinoidea (pp. 40 — 56) will recognize that in\- classification is not 

 based on the pedicellariæ alone. It is true that m\- classification of the Cidaridcc is almost exclusiveh- 

 based upon the structure of the pedicellariæ; but that is due to ni\' inabilit\' to find other characters 

 which could be used witli success. Any reader of ni\ introductor)- remarks on the Family Cidaridæ 

 will recognize that I have not omitted to take other characters into consideration, while tlie conclusion 

 of that chapter is as follows (p. 31): VVhen in the diagnoses of genera given here other features than 

 pedicellariæ and spicules have ouly beeu uientioned exceptionalh the ojjinion of conrse is not that 

 these structures should be sufficient for definitive diagnoses. It has alread\' beeu emphasized above, 

 and I shall here emphasize once more that all these structures must l)e thoroughly examined in order 

 to get the mutuai relations of the forms established. That I ha\e here onl\- treated the pedicellariæ 

 more thoroughh is a consequence of the faet that ueither m\- material nor my time has permitted 

 me to treat the other features more particularly. The system of the Cidarids cannot get its definitive 

 formulation, until all features have beeu examined in a greater number of .species (or best in all 

 species). What is gi\en here is a provisioual classification, which can scarcely be correct throughout...« 

 — Whilst I must thus decidedly protest against the accusatiou of having based m\- classification on a 

 single character, I beg to suggest to Professor Agassiz whether that would not suit the classification 

 oi\h& Echinomefridæ anå Echinidæ given in the Revision of Echini . These Families > are founded 

 exclusively on the number of pores in the ambulacral piates, all the genera with onl\- three pairs of 

 pores being included in the family Echinidce. those witli more than three pairs of pores in the famii\- 

 EchiHometridæ. And as for the impossible associations resulting from such artificial divisions according 

 to one character I might suggest to Professor Agassiz whether the placing of Hemiprdina. Phymo- 

 sorna, Echituis, Toxopiuustcs, Tripiicustes and Evechinns ( Hcliocidaris} in one subfamily, Triplechinidæ , 

 as is done in the Re\ision , does not deserve to be thus characterized. 



Professor Agassiz speaks in a \ery depreciator\- manner of the results ofmy classification, which 

 «cnlminate in such impossible associations that we are loath to follow his lead . It would have beeu 

 ver\ interesting to hear some instances of these impossible associations, but unfortunately Professor 

 Agassiz confines his examples to a few Cidarids. It scarceh- seems fair to condemu the whole of my 

 results on the evidence of a few debatable cases among the Cidarida-, the classification of which 

 family is expressh- stated to be purel\- provisioual. I should like to learn what are the impossible 

 associations in my classification of the Ecliiiiothnridcr. Echiiwiiictridfr and the E.chinidcr, the more so, 

 since it was the greater naturalness of the associations resulting from ni\ classification which were 

 to my mind a proof of its correctness. I will, however, leave it to others to compare m\- arrangement 

 of the forms included in, let ns say. the genus Strongylocintrofns or in the Famih- Echinometridæ with 

 the arrangement given in the (Revision of Echini . And upon the whole I \entnre to believe that, 

 since Professor Doderleiu has now accepted my classification of these groups in the main points, it 

 will be agreed, at least, that it cannot be so ver\- unnatural; otherwise, so careful and judicious a natu- 

 ralist, with so profound a knowledge of the whole class, would certainh' not have accepted it. 



Regarding the use of pedicellariæ in the definition of systematical characters of Echini Profes.sor 

 Agassiz agrees that it uia> l)e desirable to employ all the data possible from whatever source, 



