In Clepsine the mesoblastic bands are stated by Kowalevsky to become split into somatic and splanchnic layers, between which are placed the so-called lateral sinuses. These sinuses form, according to Whitman, a single continuous tube investing the alimentary tract; a tube which differs therefore to a very small extent from the normal body cavity of the Chætopoda. The somatic layer of mesoblast no doubt gives rise to the circular and longitudinal muscular layers of the embryo. The former is stated to appear the earliest, while the latter, as in the Oligochæta, first takes its origin on the ventral side. A delicate musculature, formed mainly of transverse but also of longitudinal fibres, would appear to be developed independently of the mesoblastic bands in Nephelis and Hirudo (Rathke, Leuckart, Robin, and Bütschli). It develops apparently from certain stellate cells which are found between the walls of the alimentary tract and the skin, and which probably correspond to the system of contractile fibres which pass from the body wall to the alimentary tract through the segmentation cavity in the larva of Chætopoda, various Vermes and Mollusca[147]. [147] According to Robin this system of muscles becomes gradually strengthened and converted into the permanent system. Rathke on the other hand states that it is provisional, and that it is replaced by the muscles developed from the mesoblastic somites. It is possible to suppose that it may really become incorporated in the latter system. The mesoblast, so far as is known, gives rise, in addition to the parts already mentioned, to the excretory organs, generative organs, vascular system, etc. _Excretory organs._ There are found in the embryo of Nephelis and Hirudo certain remarkable provisional excretory organs the origin and history of which is not yet fully made out. In Nephelis they appear as one (according to Robin, No. 364), or (according to Bütschli, No. 359) as two successive pairs of convoluted tubes on the dorsal side of the embryo, which are stated by the latter author to develop from the scattered mesoblast cells underneath the skin. At their fullest development they extend, according to Robin, from close to the head to near the ventral sucker. Each of them is U-shaped, with the open end forwards, each limb of the U being formed by two tubes united in front. No external opening has been clearly made out. Semper believed that the tubes were continuous with the three posterior vitelline cells, but this has been shewn not to be the case. Fürbringer[148] is inclined from his own researches to believe that they open laterally. They contain a clear fluid. [148] _Morphologisches Jahrbuch_, Vol. IV. p. 676. He further speaks of the tube as "feinverzweigt u. netzförmig verästelt," but whether from his own observations is not clear.