Thus the process of gill development in the Pelecypoda appears to lead up from a simple to a very complex type. In its original form, at all events in the most primitive form known to us, the gill is a series of short filaments, quite independent of one another, strung in two rows; then the filaments become longer and double back, while at the same time they begin to show signs of adhesion, as yet only superficial, to one another. In a further stage, the reflected portions become fused to the adjacent surfaces of the foot and mantle, while the interlamellar junctions serve to lock the two gill-plates together; finally, the mere ciliary junction of adjacent filaments is exchanged for intimate vascular connection, while the gill-plates as a whole become closely fused together in a similar manner. This theory of origin is strengthened by closer observation of the phenomena of a single group. Taking the Septibranchiata as an instance, we find that in _Lyonsiella_ the branchiae unite with the mantle in such a way as to form two large pallial chambers, the structure of the branchiae being preserved, and their lamellae covering the partition. A further stage is observed in _Poromya_. There, a similar partition exists, but it has become muscular, preserving, however, on each side two groups of branchial lamellae, separated one from the other by a series of slits, which form a communication between the two pallial chambers. A further stage still is seen in _Silenia_. There the same muscular partition exists, but the branchial lamellae on either side have disappeared, the slits between the two chambers, which occur in _Poromya_, still persisting, but separated into three groups. _Cuspidaria_ represents the last stage in the development. In the ventral chamber there appears nothing at all corresponding to a branchia; the surface of the partition appears perfectly uniform, but on careful examination three little separate orifices, remains of the three groups of orifices in _Silenia_, are observed.[276] _Relation between Branchiae and Heart._--The object of the branchiae being, as has been already stated, to aerate the blood on its way to the heart, we find that the heart and the branchiae stand in very important structural relations to one another. When the branchiae are in pairs, we find that the auricles of the heart are also paired, the auricle on the right and left sides being supplied by the right and left branchiae respectively. This is the case with the Dibranchiate Cephalopods (_Argonauta_, _Octopus_, _Loligo_, etc.), the Zygobranchiate Prosobranchs (_Fissurella_, _Haliotis_), and all _Pelecypoda_. In the _Amphineura_ (_Chiton_, etc.) there are two auricles corresponding to the two sets of multiple branchiae. In the case of the Tetrabranchiate Cephalopods (_Nautilus_) there are four auricles corresponding to each of the four branchiae. Compare Fig. 79, A, B, C, D, E.