ON AN INSTANCE OF COMMENSALISM. 989 



animal forms an extra panoply, not merely thicker, but more 

 extended, especially around the shell orifice, thereby generally 

 ensuring a longer tenancy to the Crustacean. This feature of 

 extended growth is illustrated in fig. 2, where the opening of the 

 fortress is seen to be formed, not of shell, but of infolded layers 

 of the Polyzoon. Again, the Polyzoon obviously benefits; for 

 it is in alliance with a A^gorous and successful marauder, and 

 although sedentary by nature, is continually being carried to new 

 and rich pastures. 



Accordingly, the large size of the new species is not surprising. 



We return now, after this economic excursion, to the pecu- 

 liarity of the association described. Against critics who deny 

 the non-accidental nature of it, we draw attention to one njain 

 point. Setting apart those favourite Pagurine mates : Suberites 

 and the epoecal Zoantharians, there remains the fact that a 

 " special seat " is reserved to Conopeum cominenscde among the 

 hosts of sessile epoecal animals, ready to jjoptdate every available 

 spot in the tepid tropical coast- waters. In a very short lapse of 

 time a typical complex is formed. This follows from the fact 

 that among the many home-shells of adult Petr. gramdiinanus 

 only one was found naked, obviously inhabited only a short time 

 before it was caught ; intermediate gradations to full-grown 

 adult colonies are wanting. Now, what is the reason that 

 among so many rivals the tiny Polyzoon regularly wins the 

 battle ? We have seen that the other candidates may win the 

 second round if the Hermit migrates to a spot unfit for its 

 Pol3'^zoan comrade to live in. In the struggle for life the 

 champion is slain at last, A^ow at last a fair chance is given to 

 everybody ; here you have Accident playing its part, and the 

 result .... a chaotic conglomerate of sessile marine life (fig. 14). 



Since the time of Darwin, Law has taken the place of Acci- 

 dent in biological science. 



Our case is in some i^espects comparable to the well-known 

 association of Eupagurus hernhardus and Hydractinia echinata. 

 Tn this case the Pagurid inhabits the whole " littoral " region 

 up to a depth of 270 metres, but Hydractinia echinata is a 

 coastal form which does not follow the adolescent Eupagurus 

 to depths exceeding those of the centime of the North Sea. 

 At the lower limit of its occurrence the Hydractinia is often 

 replaced by Alcyonidium gelatinosum. Now, although Hydrac- 

 tinia echinata has been found in a few instances without its 

 mate Eupagurus hernhardus^ is there anybody who will seriously 

 contest the particular relation between these two organisms ? 

 Of. Eupag. puhescens, which occur in the North Sea as a 

 rule, associated with Sid^erites ficus, but exceptionally with 

 Zoanthus sp. 



So far as we know, the present case has not yet been described. 

 It is a remarkable fact that Ohevreux and Bouvier, among the 

 rich material of the ' Melita,' described 17 species of Pagurids 

 (Mem. Soc. Zool. France, v. 1892), but only noticed " coquille 



