424 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



GENERAL CONCLUSIONS IN KEfiARD TO THE OILLS. 



1. The epithelium of the ventral edge of the gill plates in forms like N^ucnla, 

 YoJdia, and Solenomt/a, and of the onter edge of the filament in all other lamelli- 

 branchs — morjihologieally the same thing — is speeialized by the possession of 

 "ehitin," large ciliated cells, and gland cells. This of course does not apply to the 

 deep-sea forms with a sej^tum and no gill. This probably served in the ancestors of 

 the groui) for the procuring of food as well as for aeration. However large the palps 

 may be, they are certainly not sufiicient in Yoldia \\)y obtaining food. The gills 

 here, and i^robably in other forms with plate gills, are extremely active in collecting- 

 food. Besides Yoldia, which I have observed, tliis function is evident in the others, 

 from the presence of gland cells. 



2. The single row of latero-ft'ontal or straining cells, described by Peck in Anodon, 

 and the two single cell rows of JDreissena, are not ]>resent in any (►f the marine forms 

 which I have examined, with the exception of Mytilvs. Here there is a single row of 

 straining cells, and interiorly fi'om these a row of many closely crowded, ciliated cells, 

 which I have called the cells of the compound straining line. This arrangenn-nt. 1 

 believe, will be shown to be the most usual one in lamellibranch gills. 



There may be one or two of these lines on each side of a filament (including the 

 simple lines of straining cells when they occur). I have never seen more than two. 

 In Solenomya and Anomia there seems to be no such specialization, all the cells of 

 the outer edge of the filament being of much the same size and eciually ciliated. 



I believe the function of these cells to be that of preventing foreign bodies in the 

 water fi-om entering between the plates or filaments of the gill, by means of their large 

 cilia, while allowing currents of water to do so, 



o. Gland cells were present at the sides of the frontal region, from one to three 

 appearing in a section on either side. The forms in which thej' were seen in this posi- 

 tion were iSoIenomya, Area, Mytilns, Pecten, Anodon, Vevns, Mya, Anomia, and Ostrea. 

 I was not certain about the matter in Yoldia, though I believe that I have seen them 

 in the gill of that form; and the fact that I haA^e seen carmine grains firmly cemented 

 into considerable masses on the gill of the living animal confirms this belief. 



In Mytihis they generally appeared innnediately interior to the outer row of latero- 

 frontal cells. Lankester (No. 9) has described gland cells as occurring near the inner 

 (Mlge of the filament in Ostrea, edulis ([uite as constantly as at the sides of the fi'ontal 

 region. In Ostrea virgimana I have found them j)resent in this region in certain 

 s]>ecimens, but not in all. When they occur they seem to be always nearly spherical, 

 those outward being constantly elongated. 



In some instances, gland cells occur on the inner edge of the filament, most 

 noticeably in Anomia^ where they are extremely abundant, very much more so than in 

 the outer edge (Fig. 85). 



4. I believe that an endothelial lining of the blood cavity of the filament ov jdate 

 between the chitinous layers will be demonstrated in the majority, at least, of the 

 Lamellibranchiata. I am quite certain of having observed it in Yoldia, Pecten, Myti- 

 lns, Anodon, Venus andi)robably in Area and Ostrea. It is especially well shown in 

 Pecten and Yoldia. 



