POLVCH^ETA-BENHAM. 49 



dorsum and may even interdigitate with those of the opposite side (pi. XX VI I, fig. 3). 

 He says further (p. 444) that although they are not present in all individuals, he has 

 found them in both small and large specimens, and he suggests that their presence may 

 bear some relation to sexual maturity. I have examined several parapods taken from 

 both large and small specimens with the especial object of finding these fine capillary 

 bristles. Occasionally some of the chaetae may be seen edgewise and so appear thinner 

 than when seen on the flat surface, and frills are then seen to project from both edges 

 giving an appearance somewhat like Ehlers's figure. But I do not find such difference 

 in length as he found. Gravier does not mention their occurrence in his specimens, 

 and on the presumption that we are dealing with the same species, this is the only 

 feature in which ours really differ from those examined by Ehlers. I may add that 

 Mclntosh does not mention such bristles in his account of L. crosetensis. 



So far, then, as the present specimens are concerned, all the dorsal chaetae are 

 alike in structure, though they differ in length ; those in the lower part of the bundle 

 being about half the length of those in the upper part. 



The same difference in size exists amongst the ventral chaetae. The ventral 

 chaetae (figs. 46, 47) which in L. crosetensis, Mclntosh states are " not furnished with 

 long spines, and have a distinct sub-apical tooth;" have in the present case, as Gravier 

 has figured (pi. IV, fig. 49) certain pronounced spines or teeth amongst the upper frills, 

 which are absent in the lower frills. In some chaetae two such spines occur on one side 

 and one on the other; in other cases, two on each side. The frilled region is long, 

 consisting of about 20 frills which are discontinuous in the distal region, but become 

 continuous over the greater part. Ehlers says little about the ventral chaetae, except 

 to state that the apex is simple. Why then should he refer it to the species L. crose- 

 tensis ? 



However, in Hermadion rouchi, although most of the ventrals have a simple apex, 

 with no sign of a sub-apical tooth, there is occasionally a sub-apical " step," which seems 

 to indicate a tooth that has been worn away. And Gravier states that in some of his 

 specimens he found a tooth. 



More than one zoologist has in recent years commented upon the difficulty of 

 distinguishing between the two genera, iMgisca Malmgren and Hermadion Kinberg, as 

 well as upon the question of the distinction between them and the genus Harmotho* 

 Kinberg. Most writers accept the last genus in an extended sense as including several 

 of Malmgren's sub-genera, though Professor Mclntosh still retains most of the latter; 

 and in his splendid monograph of the British Annelids, published by the Ray Society, 

 these names are even used as generic. 



I need not discuss this matter further as Baron de St. Joseph (1888, p. 150) has 

 given the history of these names. It was Willey (1902), I believe, who first drew 

 attention to the resemblance between Lagisca and Hermadion. And Fauvel (1916) 

 has recently summarised the main points in the controversy raised by him and also 

 discussed by Gravier (1911). Fauvel concludes (p. 426) that Hermadion is distinguish- 



83892 U 



