482 MB. F. E. BEDDAED ON THE ATRIUil AND [May 16, 



in one specimen of a species recently described by me (9) as Moiv- 

 iligaster bahamensis ; in the other species investigated there was 

 no trace of any perforation of the muscular layer by the outer 

 layer of cells of the atrium : T imagine, however, that these pro- 

 longations of the cells were originally overlooked, owing to their 

 inconspicuous character ; they were especially obvious in the 

 species Moniliyaster bahamensis on account of the ducts being 

 filled with granules of secretion, which were deeply stained by 

 borax carmine. I compared this atrium with that of the Lumbri- 

 culidae and various other aquatic Oligochaeta, considering that the 

 outer glandular layer of cells in both was the modified peritoneum. 

 Benham, in the paper to which I have referred, agreed with me in 

 this matter, but held that the cells, from the very fact that they were 

 peritoneal, could not exert a glandular function ; he doubted 

 whether there was any known case of " ccelomic epithehum being 

 converted into a gland which pours its secretion to the exterior." 

 This doubt appears to me to be quite justified, but then there is 

 the undoubted fact that the processes of the cells in question do 

 pour their secretion on to the exterior through the lumen of the 

 atrium. The explanation of the difficulty seems to be that I was 

 wrong in regarding this glandular epithelium as mesoblastic at all ; 

 I have never seen it myself, but Mr. Benham has shown me 

 sections through the atrium of Moniligaster in which there was a 

 distinct, though thin layer of peritoneum outside the glandular 

 cells. The presence of this does not, however, invalidate my com- 

 parisons, though it puts them in a difierent light, for Benham 

 remarked apropos of Jihijnchelmis that he could " detect in sections 

 a delicate membrane surrounding the pear-shaped cells which is 

 no doubt the true ccelomic epithelium." 



The structure of the atrium in Moniligaster has also been 

 studied by Eosa, who is disinclined to favour any " limicohne " 

 affinities in that genus ; he principally uses the characters of a new 

 Moniligastrid, Desmogaster (14), to strengthen his position. In 

 Desmogaster the atria have the following structure : — They are 

 tubular glands slightly coiled in an S-shape ; the lumen is much 

 restricted, and is floored by a layer of cylindrical cells ; outside 

 this, the walls of the atrium are mainly muscular, the fibres being 

 disposed centrally in a circular direction and longitudinally outside : 

 among the external bundles of fibres are groups of gland-cells which 

 by means of long ducts (formed merely by a prolongation of their 

 own substance) reach the lumen of the atrium ; but these groups 

 of cells are not exposed at the surface of the atrium on account of 

 the longitudinal muscular fibres ; the entire organ is covered 

 externally by a delicate layer of peritoneum. As E-osa points out, 

 this seems to indicate a gradual moving outwards of the pear- 

 shaped cells from their primitive position as lining the lumen of 

 the atrium. My own criticism of this paper (as concerns the facts 

 about the atria), published three years ago, need not here be referred 

 to, as the data upon which it rested are not sufficient. I am now 

 quite of opinion that Eosa was perfectly right in looking upon the 



