While these changes have been taking place on the exterior, the hypoblast cells have increased in number (additional cells being probably derived from the three large vitelline spheres) and fill up in a column-like fashion a space which is bounded behind by the three vitelline spheres, and in front by the epiblast of the anterior end of the embryo. At the sides of the hypoblast the mesoblast has become established, probably as two lateral bands. The origin of the cells forming it has not yet been determined. The hypoblast cells in the succeeding stage arrange themselves round a central archenteric cavity, and at the same time rapidly increase in size and become filled with a secondary deposit of food-yolk. Shortly afterwards a mouth and thick-walled oesophagus are formed, probably from an epiblastic invagination. The mesoblast now forms two curved lateral bands at the two sides of the body, equivalent to the mesoblastic bands of the Chætopoda. The three vitelline spheres, still largely uncovered by the epiblast, lie at the posterior end of the body. The embryo grows rapidly, especially anteriorly, and the three vitelline spheres become covered by a layer of flattened epiblast cells. Around the oesophagus a cavity traversed by muscular fibres is established. Elsewhere there is no trace of such a cavity. The cephalic region becomes ciliated, and the dorsal part of it, which represents a rudimentary præ-oral lobe, is especially prominent. The cilia of the oral region are continued into the lumen of the oesophagus, and at a later period are prolonged, as in Lumbricus, along the median line of the ventral surface. The mesoblastic bands would seem from Bütschli's observations, which receive confirmation from Kleinenberg's researches on Lumbricus, to be prolonged dorsally to the oesophagus into the cephalic region. Posteriorly they abut on the large vitelline spheres, which were supposed by Kowalevsky to give origin to them, and to play the same part as the large mesoblasts in Lumbricus. It has already been shewn that the function of the large cells in Lumbricus has been exaggerated, and Bütschli denies to them in Nephelis any share in the production of the mesoblast. It seems in fact probable that they are homologous with the three vitelline spheres of Clepsine; and that their primitive function is to give origin to the hypoblast. They are visible for a long time at the hind end of the embryo, but eventually break up into smaller cells, the fate of which is unknown. The embryo of Hirudo would appear from the researches of Robin to develop in nearly the same way as that of Nephelis. The anterior part is not however ciliated. The three large posterior cells disappear relatively early. _General history of the larva._