This progressive fusion of the mantle edges may be taken as indicating definite stages in the development of the Pelecypoda. A perfectly free mantle edge, joined at no point with the edge of the adjacent mantle, occurs in _Nucula_, _Arca_, _Anomia_, and _Trigonia_ (see Fig. 80, A, B). Here there is nothing in the nature of a siphon, either anal or branchial; in other words, no contrivance exists to prevent the spent water which has passed over the branchiae from becoming mixed with the fresh water which is to reach them. When the mantle edges are fused at _one point only_, this is invariably on the middle part of the posterior side, thus separating off an anal opening which may become prolonged into a tube-like form. At the same time the adjacent underlying portions of the mantle edges draw together, without actually coalescing, to form an opening for the incurrent stream of water, the rudiments of the ‘branchial siphon’ (Fig. 80, C). This is the case with most _Mytilidae_ (see Fig. 75) with _Cardita_, _Astarte_, and _Pisidium_. In the next stage the branchial opening is separated off by the concrescence of the mantle edges beneath it, and we have the mantle united _in two places_, thus forming three openings, the ventral of which is the opening for the protrusion of the foot (Fig. 80, D). This is the case in _Yoldia_, _Leda_, the majority of the Eulamellibranchiata (_e.g._ _Lucina_, _Cyrena_, _Donax_, _Psammobia_, _Tellina_, _Venus_, _Cardium_, _Mactra_), and all Septibranchiata. In _Chama_ and _Tridacna_ the fused portions of the mantle become more extended, and in _Pholas_, _Xylophaga_, _Teredo_, _Pandora_, and _Lyonsia_ this concrescence takes place over the greater length of the whole mantle edge, so that the mantle may be regarded as closed, with the exception of the three apertures for the foot and the two siphons (Fig. 80, E). In certain genera there occurs, besides these three apertures, a fourth, in the line of junction between the pedal and branchial orifices. It appears probable that this fourth orifice (which has been regarded by some as an inlet for water when the siphons are retracted), stands in relation to the byssal apparatus (Fig. 80, F). In _Lyonsia_, for instance, a thick byssus protrudes through the orifice, which is large and open. In _Solen_, _Lutraria_, _Glycimeris_, _Cochlodesma_, _Thracia_, _Aspergillum_, and a few more genera, which have no byssus, the orifice is very small and narrow. It is possible that in these latter cases, the byssal apparatus having become atrophied, the orifice has been correspondingly reduced in size.[281]