Ihitt'sh /^y>aci«5 0/ Siphonostoma. 193 



of intermediate length arc iiitcrineJiatc iit character between 

 the two. 



Specimens of GO-70 inilliin. in length ha 1 4()-t5 aegm^iiti. 

 This does not agree with St. Joseph's statemcsnt tor S. diplo- 

 chaitonj but Bles *, in speaking of this species (Naples 

 specimens) says 40-50 segments, so that there must obviously 

 be variation in this respect. The gills in these full-grown 

 specimens numbered about 100, the cephalic bristles 2)0 or 

 more ; but in the structure of notopodium and neurop )dium 

 they showed less close agreement wiih St. Joseph's speeiinons. 

 The notonodium bore usually 10-12 long annulated bristles, 

 but in adtlitioji there were about 10-13 other short bristles 

 buried in the substance of the papilla, and hardly, if at 

 all, visible externally (see I'l. IV. tig. 1). The distance of 

 the annuli apart varieil from 0"02(j millim. to 071 millim. 

 in the cefdialic bristles, and from 0"0079 millim. to 02 { 

 millim. in the other notopodial bristles (see tig. 3). The 

 range of variation in this respect is thus very wide. As to 

 the hooks of the neuropodium, these did not exceed 3 in 

 number (in place of 4-5), and in many instances 1 or 2 only 

 were present; but it may be noticed that the hooks fall out 

 very readily. The annulation was often indistinct throughout 

 a portion at least of the hook (see fig. 4), and in no case did I 

 succeed in Knding any one hook which displayed all the 

 characters described by St. Joseph, lie describes his speci- 

 mens of S. diplocha'ilos as having less distinctly recurved 

 hooks than those of S. ajfiius ; but my specimens showed no 

 such distinction. Embe ided in the neuropodial pajiillte and 

 surrounding the base of the hooks there are, as in -b'. aj/inisy 

 short annulated bristles whose slender tips only project from 

 the surface of the papilla (see tig. 2). In the Millport speci- 

 mens instead of 6 of these there were usually about 13 : of 

 these about 7 actually projected from tiie papilla ; the other G, 

 though identical in structure, lay entirely within the substance 

 of the papilla (see lig. 2). It seems impossible to doubt the 

 identity of these Millport specimens with Siplionostoma diplo- 

 chidtos of the Mediterranean ; but they ditfer especially in the 

 diminished number of ventral hooks and in the presence of 

 small bristles in the notopodial papilla, in addition to the 

 typical number. 



The sn)all Millport specimens differed in several respects 

 from the large. Specimens of 25 millim. in length had 

 34-35 segments, but one of 38 millim. had 43 segments. In 

 specimens of 25-40 millim. the number of cej)halic bristles 



♦ Brit. Assoc. Report, 1891. 

 Ann. (1; Mag. X. Hist. Ser. 7. I o/. v. 13 



