78 



mesoblastic somite, which have begun to form the muscle plate M.P., 

 are still, in this section, continuous with the cells of the body wall. 

 In Fig. 4B the cells of the muscle plate M.P. turn back at o and the 

 segmental tubule is seen to pass from the body cavity to the Wolf- 

 fian duct. 



Appearances similar to those seen in fig. 4B have led some ob- 

 servers to speak of these tubules as if they arose first in connection 

 with the somites and then detached themselves from the somites, to 

 join the Wolffian duct subsequently. In point of fact the tubule 

 can from the very first be traced as a solid rod of ceUs, right up to 

 the Wolffian duct. When a section shews a tubule with apparently 

 a free end detached from the duct, this is because the tubules bend, 

 and are often cut across at a bend: on tracing out these tubules in 

 other sections of the series they can invariably be traced to the duct. 



Formation of the Malpighian bodies. 



This I have and studied most carefully in the embryo of the 

 rabbit. After the segmental tubule has ceased to open into the body 

 cavity it describes a sharp bend upon itself (fig. 5 ST. The end of 

 this tube broadens out into the shape of a hollow spoon, and moulds 

 itself upon this bend. Between the hollow spoon, which forms the 

 Bowman's capsule, and the bent tubule, vessels enter from the aorta 

 and from the cardinal vein to form an intervening layer, the commen- 

 cing glomerulus. This latter grows, and becomes globular, greatly 

 increasing in size, while the Bowman's capsule grows around it. There 

 is no stage during which the Bowman's capsule is saccular, previous 

 to an invagination by the glomerulus. No such process takes place 

 either in the Wolffian body or the kidney. The Bowman's capsule 

 is like a double walled cap growing pari passu with the head it covers. 



There are no secondary tubules in the chick and 

 rabbit. 



Secondary tubules, branching from those first formed have fre- 

 quently been described, but they do not appear to exist in higher 

 forms. In the rabbits embryo of 14 days fig. 6 the tubules and the 

 Malpighian bodies appear at first sight to have increased in numbers ; 

 often five or six Malpighian bodies may be seen in the same section. 

 This is due, in my opinion to the great bending of the embryo at this 

 stage, whereby the sections are all oblique, and more somites are 

 involved in each section. It is also due to the great increase in size, 

 and consequent closer packing together of the tubules, and of the 

 Malpighian bodies. In any case there are no branching tubules to 



