These fibers, which are at right angles to the 

 main fibers extending from one valve to the other, 

 are found near the attachment of the adductor 

 to the valves (fig. 149). Their position suggests 

 that they act as braces by bringing together and 

 tiglitening the principal bundles. 



250 



Figure 149. — Longitudinal section of a piece of partially 

 relaxed muscle near the attachment to the valve (right 

 side). Note band of muscles at right angle to the main 

 fibers. Kahle, hematoxylin-eosin. 



ATTACHMENT TO SHELL 



Tlie adductor muscle of C. virginica is fastened 

 .so strongly to the sliell that when tlie valves are 

 forced apart the nmscle lireaks in the middle 

 instead of tearing from the shell. The adhesion 

 sometimes withstands a pulling force of 10 kg. 

 (22 pounds). On the other hand, the connection 

 between the muscle and the shell can be weakened 

 or completely destroyed by applying lieat to the 

 shell o\"er tlie area of the muscle scar. This con- 

 nection is smooth and glossy. 



Brilck (1914) found that in the shells of AnodonUi 



and Cyclas the muscles are fastened by means of 

 a specialized layer of cells which he called holding 

 or adhesive epithelium ("haft epitlielium"). 



Hubendick (19.58) used both electron and light 

 microscopy to demonstrate the presence of ad- 

 hesive epithelium in the areas of attachment of 

 the muscles of the fresh-water snail Acroloxus 

 lacustris (Maxwell). The surface of the cells 

 has a dense brush border of minute microvilli 

 which are transversed by very thin cytoplasmic 

 fibrils originating in the base of the cell. The 

 epithelial cells are fastened to the underlying 

 connective tissue by the evaginations which 

 extend into the base of the cells. Since the 

 muscles used by Hubendick were fixed in osmic 

 acid, which resulted in their detachment from 

 the shell, the electron micrographs published in 

 his paper do not show the actual connection 

 between the microvilli and shell material. The 

 shell surface over the area of the attachment has, 

 however, small depressions into which fit the tops 

 of the microvilli. It is, therefore, likely that in 

 Acroloxus the adhesion of the muscle is accom- 

 plished in this manner. 



The holding epithelium of C. virginica can be 

 seen on transverse sections of decalcified shell and 

 muscle preparations. Individual cell boundaries 

 are indistinct, but the position of each cell is 

 clearly marked by a large round nucleus (fig. 150). 

 Fine strands resembling those described by 

 Hubendick originate in the base of the cells and 

 terminate at their surfaces. They are not visible 

 at low power but can be seen under oil immersion. 

 The holding epitheliiun of the oyster is a modifi- 

 cation of the surface epithelium of tlie mantle; 

 the transition from one type to another can be 

 seen in the areas adjacent to the muscle attach- 

 ment (fig. 151). The holding epithelium of C 

 virginica secretes an organic film of about 2 yu 

 in thickness that consists of adhesive material by 

 whicii the muscle fibers are attached to the shell. 

 The chemical nature of this film was not deter- 

 mined, but staining properties suggested the 

 presence of collagen. Since it is known that un- 

 der proper conditions collagen is digested by 

 coUagenase, I made a series of experiments at 

 Woods Hole to determine the effect of this en- 

 zyme on the attachment of muscles. Small 

 amounts of phosphate buffer solution (pH 8.4) 

 containing 1 mg. of collagenase per ml. were 

 injected into adductor muscles through holes 



160 



FISH AND WILDLIFE SERVICE 



