ECHINOIDEA. II. 67 



spines. In tliese characters P. inira)ida agrees with P. lagwicula. and Agassiz ( Challenger -Echini 

 p. 137) is certainly right in stating that this species is closely allied to P. iiiirandai. Also the P. Tanneri 

 is regarded by Agassiz as closely related to P.lagunaila; it is, however, not clear from his othervvise 

 (regarding the structure of the test) very elaborate description and figiires of this species, whether it 

 agrees with lagiiiicula (and iiiirtn/da) in the sliape of the spines and the developnient of the front ten- 

 tacles. Of the spines it is only said : the primår}' radioles on the flanks of the test are also longer, 

 while in /-". lagiincula and I', iii/rdi/dn they are somewhat spathiform (Pan. Deep-Sea Ech. p. 132). The 

 front tube-feet are not mentioned at all. Having received a specimen of /\ Tainuri from the U. S. 

 National Museum I can state that the .spines are not widened towards the point, whereas the frontal 

 tube-feet are really rather large and conspicuous The pedicellariæ do not afford any proof of a close 

 relationship between P. Tanneri and lagunctila. In the former species I ha\e found only rostrate 

 pedicellariæ with rather .slender valves (PI. XI. Fig. 11) and small tridentate pedicellariæ of the same 

 form as in P. Jcffreysi. 



In /-". lagunciila (examined in the British Museum) I have found (in a specimen from St. 232) 

 globiferous pedicellariæ with the valves ending in two or three long teetli, resembling closel\' those 

 of P. carinafa (comp. PI. XI. Figs. 16, 22), ophicephalous pedicellariæ with rather elongated, slender 

 valves (PI. XI. Fig. 12) — (differing considerably from those figured in the Chall. -Ech. PI. XLIII. 18 

 — 19 under the name of < Clypeastroid-like - pedicellariæ, so much, indeed, that they can scarcely be- 

 long to the same species) — and two forms of tridentate pedicellariæ, viz. the usual small form, which, 

 however, here occurs also with the apophysis continuing into the outer edges of the blade, and a 

 larger form with long and slender valves with the blade almost flat (PL XI. Fig. 33), the outer edge 

 very finely serrate. (This form differs so much from the pedicellariæ of the other species that it ma\- 

 perhaps be suggested not to belong reall\- to this species). Of rostrate pedicellariæ I have found onl\- 

 one small specimen, which does not differ essentially from those of /'. lanniri. vSmall spicules, in the 

 shape of fenestrated piates are found in the large frontal tube-feet. 



The form figured in the Challenger -Echinoidea PI. XXXI. 7 — 11 and mentioned (p. 138) as 

 younger specimens of P. lagiincula showing considerable variation in the outline can hardl}- be the 

 same species as that figured in the same Plate, Figs. i — 6, which must be taken as the type of the 

 species. The latter .specimen vvas 22""", that represented in the figure 7 — 11 was 12'"'" in length. It 

 seems hardly conceivable how so great a difference in the shape of the test could be due merely to 

 changes during growth, and a growth onl\- from 12 to 22'"™ in length. This is made even more un- 

 likeh- vvhen we learn ( Chall. -Ech. p. 138) that some of the specimens with narrow anal snout char- 

 acteristic of the smaller .specimens measuring from 12 — 16""" were nearly 19™'" in length . The con- 

 clusion seems quite inevitable that this form with the narrow anal snout is a distinct species, which 

 will perhaps prove identical with P. Taiinrri. The material preserved in the British Museum does not 

 give the solution of the question, since no specimen is found which can with certainty be recognized 

 as belongiug to the narrow type (« Chall. »-Ech. PI. XXXI. 7 — 11). Specimens of the broad type, the real 

 P. lagiincula are 2:>reserved from St. 232 and St. 191 (the latter are badly crushed, but can, however, be 

 recognized as belonging to this form); from St. 169 small fragments only are preserved, which cannot 

 be recognized as belonging to either of the forms, and the same is the case with the anterior ends 



9* 



