332 Ealph S. Lillie 



partitiou, extending from the ventral blood-vesscl to the ventro- 

 hiteral body-wall. It usually becomes incomplete at the second or 

 third soniiie in advance of the growing regiou (Plates 22 and 23, 

 Figs. S — 10, 29, 30), apparently as a result of the thinning and 

 evcntual dii^ruption of its dorsal portions. The conditions that 

 lead to this result seem to depeud in part upon the progressive 

 inerease in the transverse diameter of this region of the body, 

 associated, as above described, with au inerease in the extent of the 

 body-wall. A further reference to Piate 23, Figs. 16 — 21, which 

 represent sections through successive somites of a single larva, will 

 show that the inerease in transverse diameter is dependent chiefly 

 upon the extension of the thin-walled dorsal and lateral regions of 

 the body-wall. It seems, therefore, probable, as above suggested, 

 that the septa first become incomplete in the dorsal region largely 

 in consequence of the undue stretchiug to which they are there 

 subjected, which has caused here a thinning and eventually a dis- 

 ruption and disappearance of the septa. On the other band, their 

 ventral portions persist for a very considerable period, and play an 

 iraportant part in later development in connection with the nephridia 

 and ncphrostomial blood-vessels. 



ic) Formation of the Nephridia. 



I shall now j)roceed to a description of the histogenetic pro- 

 cesses by which the nephridia are differentiated out of the originally 

 uniform somatopleuric layer of the mesoblast. In the region imme- 

 diately adjoiniug the pole-cells this layer, as described above, is 

 comi)Osed of a continuous uucleated protoplasm without cell-boun- 

 daries and without visible diöerentiation; at a slightly more anterior 

 l)Osition, however, where the somatopleure passes iuto continuity 

 with the newly forming septum, thc dififerentiation of the longitudinal 

 muscle-fibres of the body-wall is already in progress. These 

 structures make their first appearance at a very early period, in 

 tliat portion of thc mesoblast which is in immediate contact with 

 the ectoderm of the body-wall (Piate 23, Figs. 13, 14, 21, 24). The 

 libres first become visible in two ventro-lateral areas on either side 

 of the ventral nerve-cord, which is just beginning its appearance 

 (//.c); and at a slightly later period thcy appear also in two dorso- 

 lateral areas on either side of the dorsal mescntery. Along the 

 line of insertion of the dorsal setae muscle-fibres do not at first 

 aj)pear. 



