Xatural IlisfovT/ of East Fininark. 573 



characters, sometimes also generic ones. . . It may, perhaps, 

 seem unreasonable to lay so much stress, as is done here, on 

 so minute features as the pedicellarire — to use them for the 

 characterizing of as well species as genera and families. But 

 when it proves to be a real fact that these minute features 

 give excellent constant characters, it may be taken to be 

 reasonable to use them without regard to their being small 

 or large . . . The supposition by Stewart that by the exami- 

 nation of the pedicellariae &c. we might find a closer relation 

 between forms not otherwise regarded as related, has been 

 amply justified by these researches, even to so high a degree 

 that the classification hitherto used proves to be quite a 

 failure (with regards to the groups treated of here). A good 

 proof of the correctness of the new classification given here, 

 which has been found especially by the examination of the 

 pedicellarire, is found in the fact that forms with the same 

 kind of pedicellaricC also agree in other important respects." 

 The avicularia have been little used in the classification 

 of Polyzoa, but I am satisfied that they are destined to play 

 a far more prominent part in the future. In some genera 

 Hiucks made use of them with good results ; in others he 

 disregarded them altogether and left genera (e. g. Membrani- 

 pora, Schizuporella, Mucronella, and Lepralia) to contain a 

 most miscellaneous assemblage of species. Busk, in his 

 'Challenger' Report, used them with satisfactory result, espe- 

 cially as applied to the very difficult genus Cellepora. But the 

 following sentences from the paper by Waters, " Observations 

 onthe^Mcmbraniporidffi''^ (Journ. Linn. Soc, Zool. vol. xxvi. 

 1898, pp. 655-657) relate to a more minute point among his 

 '' jNIembraniporida^.-'^ He says: "the avicularium only ex- 

 ceptionally has a complete bar." Then writing of an aberrant 

 group (the genus Chaperia, Jullieu) he says : " Kirkpatrick 

 refers Chaperia acanthina, Q. & G., to Lepralia, but in 

 Chaperia the avicularia have not a complete bar; whereas 

 in all the Leprali(R I have examined the bar is complete, and 

 the muscular attachment of Lepjralia is not quite similar." 

 I have confirmed Waters-'s statement as to the incomplete bar 

 in the avicularia of Membranipjora in the folloAving species : 

 jlustroides, lineata, craticula, auriia, Bumerillii, unicornis, 

 armifera, Sophice, nigrans, tenuirostris, granulifera, trifolium, 

 and Flemingii. But the bar is incomplete also in other 

 genera, e. g. Lepralia uitiila, Reptadeonella viotacea, Cribilina 

 punctata, innominata, and radiata, and Mucronella (?) puvo- 

 nella; whileit is complete in Cribrilina fiyularis, Chorizopora 

 Brongniartii, Microporella ciliata, Schizoporella unicornis, 

 linearis, and other species of the genus, Smittia trispinosa, 



