TRANSACTIONS OF SECTION D. 789 



mirabiUs, but forminnf a single series. In the original description (op. cit., part 2, 

 vol. xiv. 1886, page 331, and pi. xliii. fig. 10) I wrote as follows : ' At the base of 

 the atrial siphon, where the invaginated layer of test ends, there is a slight ridge 

 which bears a series of small tentacles projecting freely into the peribranchial 

 cavity. These atrial tentacles are much smaller than the ordinary or branchial 

 tentacles, and there are only twelve of them. The position of the atrial tentacles 

 in relation to the atrial siphon corresponds exactly' to the position of the branchial 

 tentacles at the base of tbe branchial sipliou, but their use at the entrance to the 

 peribranchial cavity is not obvious. It has been observed in some simple ascidians 

 that the current of water which usually flows in at the branchial aperture and out 

 at the atrial is occasionally reversed for a short period, the atrial aperture be- 

 coming inhalent. Possibly in the present species this habit may have become so 

 marked as to have favoured the development of a circle of atrial tentacles, which 

 would act as tactile organs waving in the current of water entering the animal.' 



During the last few years I have found similar atrial tentacles in at least three 

 new species of the compound (?) ascidian genus Chorizoconnus, viz., Ch. sydney- 

 ensis, Ch. leumphceus, and Ch. subfuscus, all from Australia. In each case they 

 form a single circlet, as in Goodsiria placenta, and there are about twenty tentacles. 

 They are briefly referred to in my ' Revised Classification of the Tunicata ' (1891), 

 at page 636, and will be figured in the forthcoming ' Catalogue of Tunicata in the 

 Australian Museum.' And, lastly, Julin has made the interesting discovery that 

 ntrial tentacles are also present in Styelopsis grossidaria. 



I have queried above the genera Goodsiria and Chorizocormus as being com- 

 pound ascidians because they belong to the family Polystyelidte, in regard to 

 which it must be considered still doubtful whether the masses of ascidiozooids are 

 true colonies. But although they may be colonial forms now, there can be no 

 doubt that phylogenetically the Polystyelidre are closely related to the sub-family 

 Styelinfe of tbe Cynthidaj, the sub-family to which both Bathyoncus and Stye- 

 lopsis belong. So we arrive at the interesting conclusion that the four genera in 

 which up to now atrial tentacles have been noticed, although differing widely 

 from one another in appearance, structure, and habitat, are yet phylogenetically 

 rather closely related. 1 think it not unlikely that atrial tentacles will be found, 

 if looked for, in other members of the groups Styelinfe and Polystyelida3. 



Another point : it is an interesting fact, and may have some significance, that — 

 putting aside Bathyoncus mirahilis, in regard to the conditions of life of which we 

 know nothing — all tbe five species in which atrial tentacles have as yet been 

 demonstrated form either colonies or aggregations, i.e., they have numbers of small 

 individuals or ascidiozooids massed together. It is quite conceivable that, under 

 these crowded conditions, it may be some advantage to tbe animals to have the 

 power (to return to the suggestion I made in the Challenyer Report) of frequently 

 reversing the current of water or of using the atrial for a time as the inhalent 

 aperture — possibly, for example, because of being so placed amongst neighbours 

 that the atrial siphon is able to draw upon a purer supply of water — and in any 

 such case the advantage of having the entrance to the peribranchial cavity pro- 

 vided with a circlet of tentacles is obvious. 



Possibly Julin, in his recent investigations on Styelopsis yrossidaria, has already 

 solved the problem of the meaning <ind use of tbe atrial tentacles, and may be able 

 either to support or demolish the above suggestion. In any case I shall be much 

 interested to learn his views on the matter. 



11. Notes on Lampreys and Sags. By J. Beard, M.Sc, Ph.D. 



Two small specimens of Myxine ylutinosa were described. The one measured 

 G'5 centimetres, the other 15 centimetres. It was stated that an Ammoccete stage 

 appeared to be unrepresented in Myxine, but that the individual underwent a 

 metamorphosis, as was proved by the 'diphyodont' dentition. 



In the smaller specimen the external characters were those of the adult, but the 

 internal anatomy revealed among other things several rows of teeth in both roof 



