

378 FUNAFUTI ATOLL. 



The presence of accessory tentacles in the basal group of cor- 

 midia, appears to have hitherto been overlooked in the genus 



Prof. E. Haeckel, in the " Challenger " Report,* remarks that 

 " The smaller basal group, at the posterior or distal end of the 

 trunk, produces merely a series of small siphons and palpons, 

 placed before the protosiphon and is provided with a single ten- 

 tacle only ; it always remains sterile and never produces gono- 

 phores." And again, on page 344: "The primary tentacle which 

 belongs to the protosiphon, remains either as the single tentacle 

 of the basal cormidium, or it is afterwards lost ; but I have never 

 seen secondary tentacles developed in this distal group." 



From the above remarks it seems clear that the secondary ten- 

 tacles occurring in the basal group of cormidia have escaped notice. 

 This may be due to imperfect or ill-preserved specimens from which 

 some of the species have been described. 



In living or well-preserved examples of either Physalia utriculus 

 or P. megalista, the basal tentacles are very conspicuous and may 

 be easily seen by the unaided eye. In badly preserved specimens, 

 in which the tentacles are generally more or less contracted, they 

 are not so evident ; they can, however, be readily distinguished 

 with a hand lens of moderate power. 



With a view of rendering it less difficult to separate the two 

 Pacific species, I have carefully dissected and measured a series 

 of specimens of each. The results are embodied in the accompany- 

 ing tables. 



In dissecting the specimens, I began by isolating the anterior 

 cormidia, and afterwards snipping off the entire bunch of append- 

 ages without rupturing the pneumatophore. The siphons, tentacles, 

 and gonodendria were then separated and counted. The palpons 

 have not been taken into account. 



In the first twelve enumerated in the table, the siphons of the 

 basal groups have not been noted. In the last six, the whole of 

 the cormidial appendages palpons excepted both ventral and 

 basal have been enumerated. The gonodendria were counted 

 according to age ; thus, in some cases, as many as four occur in 

 one cormidia, all being in a different state of development. In 

 the larger examples of P. utriculus, it often proved difficult to 

 determine whether the last (sixth) ventral cormidium should be 

 regarded as one cluster or two ; frequently there is a clear space 

 on each side, indicating two pedicels, but the dividing line is not 

 continued through the centre. 



* Haeckel " Challenger " Keport Zool. xviii., p. 311. 



