MOEPHOLOGY OF LAMELLIBRANCHIATE MOLLUSKS. 415 



another, and the same number on the other side, opposite these, may contract them- 

 selves apparently from every direction, thus giving the appearance of a deep groove 

 running entirely around the gill. Ni>w this contracted zone begins to move along the 

 length of the gill and it may move in either direction. In this wave of contraction but 

 three or four plates on either side are ever concerned at one time. These waves are often 

 single, and at times several may follow one another in succession. 



Still another wave contraction may often be noticed on the ventral sides of the 

 phites. The ventral side of a single plate, or at most two plates, is affected at one 

 time. These waves occur independently on either side of the gill. A single i)late 

 bends a certain region of its ventral surface forward or backward, so as to separate 

 this region from one of the neighboring plates and bring it close to the plate on 

 the other side, either before or behind it, as the case may be. This latter j)late quickly 

 bends in the same way, the first one assuming its original i^osition and then the 

 succeeding plates, thus causing a wave of this bending to run along the length of 

 the gill. These waves may run either forward or backward. Single plates may con- 

 tract slightly independently. 



If a gill be dissected out from the body, these contractions still continue to take 

 Xjlace whenever it is touched. The action of the cilia is so x>owerful that the entire 

 gill is made to move about in the water. 



Collection of food. — Mitsukuri came to the conckision that this j)rimitive form of 

 gill was concerned only in the aeration of the blood, and that it was i^robably not 

 concerned in the procuring of food. I was able to observe in the case of Yoldia, how- 

 ever, that not only was the function of gathering food possessed by the gills, but that 

 it was performed with amazing rapidity. Carmine particles in the water once com- 

 ing in contact with the ventral edges of the plates, having been swept there by the 

 powerfril cm-rents which these ciliated borders set up, are at once hurried along toward 

 the wide, median, ventral groove of the gill, into which they are thrown. On the 

 way to this groove they have evidently become covered by mucus secreted by gland 

 cells ; for the separate particles of carmine are soon firmly cemented together, and 

 passing along the groove anteriorly, less rapidly than on the edges of the plates, 

 though stiU at a comparatively fast rate, they are finally piled up at its anterior end 

 and gradually passed upon the surface of the palp. 



Structure. — A diagrammatic view of the ventral side of the gill is given in Fig. 78, 

 PI. xci, in which the thickened ventral edges of the plates (jp) are shown on either 

 side of the groove {gr). These edges in the living gill are only slightly bent down- 

 ward, and are not so curved in outline as represented in Fig. 81, which was drawn 

 fr'om a hardened specimen. From the groove [gr^ Fig. 78) to the point cy, the plates 

 are in no way connected with one another, and one may see entirely through the gill 

 between them, when examining from below. 



At the point cj, where the edge of the plate turns abruptly upward, occurs a cili- 

 ated junction not before described in these forms, between adjoining plates. The cilia 

 from neighboring plates interlock closely, making a comparatively substantial union. 

 This ciliated junction is not confined to the point mentioned, but extends fi'om it 

 upward on the hiteral sides of the gill for about two-thirds of the distance to the 

 upper pointed extremities of the plates, the ciliated union being confined to the lateral 

 edges of the plates. 



