48 R. HOLMAN PECK. 



then terminates as a free hook-like process facing the mantle 

 (see Plate IV, fig. 10). Similarly, from the W on the right- 

 hand side (supposing always that the letter represents the 

 section of the two gill-plates of the left side of the animal) 

 there descends from the central angle or point of lophophoral 

 attachment the filament of the outer lamella of the inner 

 gill-jilate ; bends on itself, forming the angle or apex of the 

 inner gill-plate, and ascends as the filament of the inner 

 lamella of the inner gill-plate, to terminate freely in a hook- 

 like process facing the foot. Between each lamella of each 

 gill-plate — that is to say, between the descending and ascend- 

 ing portion of each filament or limb of the W — is a space, 

 the interlamellar space. 



The W-like pairs of filaments succeed one another 

 in a closely set series as we pass in the antero-pos- 

 terior direction along the lophophoral line. Each w-like 

 pair of filaments is separated from its successor by a narrow 

 space or interval called the interfilamentar space. 



Each filament in Mytilus is a hollow tube precisely similar 

 (excepting for its reflection on itself) to the filaments on the 

 lophophor of a Polyzoon (Tentaculibranch) or to those on the 

 so-called arms of a Terebratula (Spirobranch). Each fila- 

 ment is clothed externally with a ciliated epithelium and has 

 its inner wall strengthened by a chitin-like deposit. The 

 cavity of the filament is, as in Tentaculibranchs and Spiro- 

 branchs, continuous at its point of origin in the lophophoral 

 ridge with the body-cavity, coelom or blood -lymph space, a 

 special tract or " blood-vessel " being differentiated in con- 

 nection with the lophophoral ridge. The tissue which forms 

 the internal walls of the filaments is mesoblastic tissue of 

 the primitive kind common in Mollusca and Vermes. 



The course of the blood in the gill-filaments of Mytilus is 

 not easy to determine. A great deal too much stress appears 

 to have been laid on this question in connection with the 

 gill of Anodon and other Lamellibranchs. It appears to be 

 quite certain that the 7nai7i function of these organs^ is not 

 respiratory but accessory to alimentation. We are likely 

 enough to go astray if we seek for an elaborate afferent and 

 efferent system of branchial vessels in an organ which chiefly 

 serves the purpose of producing currents of water and in 

 animals where the tissue oxidation is exceedingly sluggish, 

 and the blood-lymph plasma of so lowly organised a character 

 as to be freely discharged without injury to its owner, or 

 diluted with large quantities of introduced water without 

 affecting its physiological activity. 



* See A-lder and Hancock, * Annals and Mag. Nat. Hist.,' 1856, 



