Resjnration in Invertebrate Animals. 249 



has never arrested the curiosity of the anatomist. In Mytilus 

 and in Mytilus only, Dr. Sharpey fi*^-ures correctly the manner 

 in which, at the distal margin of the lamella, the bars of the 

 upper become continuous with those of the lower lamella. In 

 Mytilus the structure of the gill is almost unique (PL VIII. 

 fig. 17). The order which obtains in nearly all other genera 

 could not b:' deduced from the anatomy of the Mytilidan gill. 

 It is a singular exception. It is the rare exception only that 

 Di". Sharpey has pictured. The rule of structure remained 

 really to be discovered. If the blunt and acute edges of the 

 penknife-shaped branchial bar (fig. 20) carry each a blood- 

 channel, then each gill in Mytilus will be a double gill, for the 

 upper and lower are identically formed. If, on the contrary, 

 the blood-channel exists only at the blunt edge {a) of the blade, 

 the current travelling peripherally along the bars of the upper 

 lamella (A, fig. 17) must turn round (as shown in fig. 24 d, d) 

 at the free margin through the loop and move centrally along 

 the bars of the lower lamella (B, fig. 17). In the latter case 

 the gill would be single, in the former double. The bars of 

 the upper lamella when the gill is single carry currents moving 

 in the same direction (PL VI. fig. 2 ; PL VII. fig. 14) from one 

 border of the gill- plate to the other ; those of the lower, oppositely 

 tending currents (PL VIII. fig. 23 e, /). This point is the 

 wonder-striking feature of the branchial enginery. No writer 

 has ever given to it a single thought. Itxleserves to be further 

 elucidated. In Pholas (PL VI. figs. I & 2), GastrochcRua, Mya, 

 Tellina, Madras Cyprccay Cardium (PL VII. figs. 13 & 14), 

 Ostrea (PL Vlll. fig. 21), and probably in many other genera, 

 the inner gill is double and the outer is single. The Pandoridaj 

 and Lucinidse are families in which the outer gill is altogether 

 suppressed. In Solen, Pecten, Unio, Venus, Kellia, Area, Sec, 

 the two gills on both sides are equal in size and double in 

 structure. 



Every gill -plate, whether single or double, is composed of two 

 lamellse, between which the excurrent water-tubes (see large 

 arrows in PL Vll. fig. 13, PL VI. fig. 1, and figs. 7 & 8) are 

 situated. In the example of the double gill each lamella is the 

 scene of a double system of opposed currents of blood, since 

 the two limbs of the same looped bar lie on the same horizontal 

 plane in the same lamella (PL VII. fig. 9 c). The adjoining 

 limbs are thus alternately afi'erent and efferent, or venous and 

 arterial. Each lamella then of every double gill is a complete 

 and independent gill. Its system of circulation is distinct, and 

 totally unconnected with that of the other lamella. Neverthe- 

 less, a single lamella of a double gill is not identical in ana- 

 tomical characters, or structurally, or perhaps officially, equi- 



