99 
ON THE MICROSCOPIC STRUCTURE OF SHELLS. 
paition of a homogeneous, intercellular substance between the colls; which 
substance answers the double purpose of giving additional finiuiess to the 
cells, am) of holding them closely together. It is evidently of a horny cha- 
racter, and seems to correspond with that wJiicli forms the b^ssus in Pinm, 
which is thrown out as the periostracnra in other families. The fact that 
in Pema we occasionally find, Iteitceen the calcified layers, membranous 
laminae, consisting chiefly of homy matter interposed betwcon rounded cells 
that are widely separated from each oilier, seems to add weight to this view 
(«* hg. 12, Plate_VI. of former Keport); since it shows that the degree of 
proximity into which the cells may be brought, depends upon the amount of 
material thus interposed between them. 
III. MjftilacetP. 
69. Referring to my former Report (§ 53) for a statemeut of the reasons 
which appew to require the removal of Pinna from this family into a closer 
spprojiniation with Avicu/a and other Marf/aritaceay I have at present to 
the structure of the shell in the genera Mytiluji, ModifU/i, and Lxtho- 
to tiiese we must add Drttsfena, which, as I shaJl presently show, is 
[hstinguished from as much by the intimate structure of its shell as 
It 18 by other characters. These shells all possess a periostracum of horny 
®ciabrane, in which 1 Imve not been able to detect any distinct structure, 
ctvecn the perioatracum ami the shell a thin layer of minute cells may be 
retiucntly detected; they are very obvious, under a sufticientiy high power, 
Jii/ttjus edtifis. TJte shell itself presents two laj’ers, which may usually 
■ / distinguished their colour, as hi tlie common mussel. The 
la criial layer is frequently iridescent, and nearly always possesses something 
® e nacreous lustre. Neitlier inner nor outer layer in Afytih/x or Modiola 
presents any ve^ distinct organic struolure; and nothing but a delicate 
cinbranoua rmdniim is left when the calcareous matter has been removed 
i! 1 * the other hand, the external layer has a remavk- 
e iiDiilar structure, which strongly resembles that of dentine. The tubules 
re 0 such extreme minuteness as not to admit of accurate measureraent; 
InT I p'l however, certainly does not exceed 1—20,00()th of an inch. 
»hn * » * having the irr<‘gular reticulated arrangement which they liave been 
and Lima, the tubuli run, like those of dentine, in 
_ .”^ *^™'ghtand parallel riiirelioii, seldom branching or inosculating, but 
'’Ufi surface of the layer to the other (fig. II). I think 
k tl • I detect indications of an original cellular constitution 
cnv(.'**i • !”h*tance, analogous to those which JVof. Owen has dis« 
structure of the shell of Dreissetm polt/morphn 
to iti«T . is so different from that of Mytilm, as fully 
lavo. • separation. The internal layer is here composed of 
^ prisinatic cells (fig. 12), of whIcJi the transverse sections made by a 
skiiU^* n^* the shell are shown at a, whilst at b are seen 
bounrf*^ divided obliquely. Notwithstanding thn distiucloess of the 
to ’’‘these cells ill a section of tie shell itself, I have not been able 
Wrtif ” hidicatlons of them in the decalcified membrane, the original 
•hite'^^ them having altogether disappeared. Between the intenial 
slsoti* perioatracum, is a layer of a brownish colour, In which 
^rllular structure present themselves, especially in a section made 
JsnLpin*^*’ ** n ‘k surface, which brings into view a decided prismatic ar¬ 
ia iS •* <’orire absence of any cellular structure 
c cate residuum which is left after the action of acid on the shell. 
H 2 
