480 JIH. F. E. BEDDABD ON THE ATRIUil AXD [May 16, 



ever, there are stronger arguments than those to be derived from 

 considerations of this kind. I have pointed out that in Merjascolex 

 newcambei the atria have the compact appearance of the corre- 

 sponding organs of AcantJiodrihs, to which also they show some 

 shght approximation in outward form ; this is further shown in the 

 slight branching of the lumen of the gland. A better instance 

 still has lately come under my observation. In the remarkable 

 genus Diplocardia ' of Garman (11) the atria appear on a dissection 

 to resemble those of other AcanthodrUidse ; they are long tubular 

 structures terminating in a short muscular duct. They ha\e, how- 

 ever, a curious spotty appearance, small white rounded masses 

 being imbedded in a darker ground-substance. They are also not 

 quite so regularly tubular as they are in all other Acanthodrilidae 

 which I have dissected ; towards the junction with the muscular 

 duct the tubes are wider, and eA erywhere the margin is slightly 

 crenate. In transverse sections the reason for these divergences 

 from the normal Acanthodrilid structure was apparent : the white 

 spots are aggi-egations of the pear-shaped glandular cells, which no 

 longer form a regular investment of the lining epi! helium of the 

 tube ; and here and there the lumen has inconspicuous diverticula 

 into which the pear-shaped cells pour their secretion. The whole 

 gland in fact shows a commencement of that branching which is 

 so characteristic a feature of the atrium in the Perichcctida;. A little 

 more development of the tendency to the aggregation of the pear- 

 shaped cells and we should have an atrium exactly like that of the 

 Perichaetidae. 



If it be reasonable on a prion grounds alone to regard the 

 glandular appendices of the sperm-ducts in the Megascolicidae as 

 homologous throughout, it is at least equally reasonable to seek 

 for detailed resemblances in the atria in the different genera which 

 make up the family of the Tubificidae. At first sight there seem 

 to be considerable differences. In Tnhifex itself, and Limnodnlus 

 and some other genera identical with Tuhifex in these particu- 

 lars, the atrium is long and has the well-known " Cement-Driise " 

 attached to it ; this gland is limited in extent, though varying 

 somewhat in different genei'a, and is superadded to the hning 

 epithelium of the atrium ; Tejdovsky (3) has shown that it is 

 developed from the lining epithelium. In Tehnatodrilus (12) 

 there is a series of these outgrowths of the atrial ejJithelium, which 

 deck the entire atrium, being symmetrically arranged. Finally, in 

 my genus BrancMura (7) the atrium is closely invested by a thick 

 layer of pear-shaped cells, which are separated from the lining 

 epithelium by a muscular coat of circular fibres. Ilyodrilus appears 

 to possess an atrium of the same kind but minus the muscular 

 fibres. In Tuhifex the muscular coat exists but it is less developed 

 than in BrancJiiura; it is, moreover, deficient at the point where 

 the " Cement-Driise " is grafted on to the atrium. The only re- 

 maining genus of Tubificid which shows any differences from these 



' I am indebted to the kindness of Dr. Benham (or the opportunity of 

 examining this worm. 



