BLOOD- VESSELS OF THE SEROUS MEMBRANES. 59 



which, without joining the other capillaries of the nodule, forms a pool 

 and passes into a vein. The part to which this capillary loop belongs 

 may already be clearly recognised to be a young developing nodule. 

 Now, between this simple capillary loop and the above-mentioned 

 completely developed blood-vascular system, the most various inter- 

 mediate stages may be found ; and it becomes a question whether in 

 time a part of the afferent limb of such a capillary loop becomes an 

 artery as the growth of the nodule goes on, and whether a portion of 

 the afferent limb becomes developed into a vein. My researches on 

 this point have led to no positive result. I could find no stand-point 

 from which I might arrive at a positive conclusion, for instance, that 

 the artery of a young nodule had been developed from what was 

 originally a capillary vessel. I cannot, however, give a decidedly 

 negative answer, since one is forced to assume, partly from the com- 

 parison of nodules in different stages of development, and partly from 

 general embryology, that this really is the case. 



We now pass to the consideration of the second point, viz. the 

 development of capillary blood-vessels in the nodules. 



Absolutely the best object for study is the omentum of the rabbit. 

 Both fresh objects examined in serum, and pencilled preparations 

 slightly impregnated with silver, give excellent results. If attention 

 be given to a series of young patches, in which the branched cells of 

 the matrix are clearly recognisable, and not obscured by the presence 

 of migratory cells; and in which, moreover, only a very limited 

 number of capillaries is formed, we see immediately that the wall of 

 these capillaries is protoplasmic to a very great extent ; and we can 

 also see, when we have before us a nucleus on the upper or lower wall 

 of the capillary, that the part immediately surrounding the nucleus is 

 distinctly granular, the neighbouring portions gradually less so, until 

 finally the wall becomes hyaline. In good silvered objects we can 

 follow the endothelial markings of such capillaries of young nodules 

 only for a short distance at their commencement and termination, 

 i.e. at the parts nearest to the artery and vein. In fully-developed 

 patches the endothelial marking is distinctly to be recognised in the 

 capillaries. Accordingly, we must consider those capillaries whose 

 walls still distinctly retain their protoplasmic character as young 



