246 Dr. T. Williams on the Mechanism of Aquatic 



in parallel directions on the same horizontal plane would form a 

 stratum of bars — such is the branchial lamella. Disposed on 

 two coincident planes, one above the other, two parallel lamellae 

 would result. Between parallelly-arranged rigid bars the inter- 

 spaces would be parallel and equal — such are the intervectal* 

 water-passages of the branchiae. If traversed by cross threads 

 at frequent intervals, a long fissure would assume the form ot 

 oblong foramina (PI. VI. fig. 1 g, fig. 2 e). Such sometimes 

 are the varieties which occur in the intervectal passages. If the 

 parallel lamella? be tied together at regular points by bands 

 running with the bars, the space between the lamella? would be 

 divided into tubes. Thus are formed the interlamellar water- 

 tubes (figs. 7, 9). The picture is faithful to nature. It mirrors 

 the reality of a complex apparatus. It represents in simple 

 outline the machinery of the branchiae in the bivalve mollusk. 



The details are now neither intricate nor unintelligible, be- 

 cause the constructive idea is clear to the intellect. In all inves- 

 tigations a tangibly- grasped mental picture must forerun the 

 clear perception of the outward reality. 



A branchial bar is a tube whose sides are comparatively rigid, 

 and whose diameter is uniform (PI. VI. fig. 6 a, a). It is clothed 

 externally by a membrane, the continuation of the mantle, of 

 which the epithelium is evolved at certain regular lines into 

 cilia-bearing scales (b). The opposed sides of each bar are 

 formed of, and supported by, cartilages (a, a). If these two 

 cartilages were far removed apart, the blood-channel would be 

 broad and flat (PI. VII. fig. 15 b, b, b). These cartilages are 

 slender in the extreme in texture; they are membraniform and 

 exquisitely hyaline ; curved at the edges, they assume the figure 

 of a hollow semicylinder ; they possess just enough rigidity to 

 preserve the straightness of the bar ; they are continuous through- 

 out the whole length of the bar (PI. VIII. fig. 17). Being- 

 placed on the opposed horizontal sides of the bars (not on 

 the upper and under aspects), they must necessarily circumscribe 

 a tubular channel of unbroken continuity. The sides are not 

 perforated by openings of any description. If the transverse 

 structures (PI. VIII. fig. 22 a), afterwards to be described, be 

 vessels or blood-channels, as conceived by Mr. Hancock ana 

 some of the elder anatomists, the bore of such channels cannot 

 communicate with that of the parallel bars. The transverse 

 parts must therefore, if they be blood-channels at all, constitute 

 an independent system. But they are not so. They are con- 



* From the Latin vectis, a bar. Since it is proposed to distinguish the 

 branchial blood-channels under the name of bars, it is only consistent to 

 mark the spaces between them as intervectal, rather than as intervaseidar. 



