SPINDLE-CELL SARCOMATA. 15> 



to keep the cells together, cannot be supposed to be lacking 

 even in spindle-cell tissue ; by its aid, and owing also to the 

 harmonious coaptation alluded to above, bands of cells are 

 formed, bundles of variable thickness, which constitute the proxi- 

 mate structural elements of the spindle-cell sarcoma. The 

 bundles very rarely radiate from a single centre; more com- 

 monly we find a number of centres ; the various systems of 

 fasciculi being interwoven wdth one another in the various 

 lines indicated by the position of those points. As a general 

 rule however, we are not able to detect any definite plan in 

 accordance with which the bundles are combined to form a 

 single whole. In every section we examine (fig. 48) we find 

 some fasciculi divided longitudinally, others transversely, others- 

 again in a more or less oblique direction. An attempt has been 





Wf 



"1 



Spindle-cell sarcoma. A^essels gaping. The cellular bands 

 divided, some transversely, some longitudinally, -j^^. 



made to connect the course of the cellular bands with that of 

 vessels ; but this is altogether wrong ; for although the thicker 

 fiisciculi invariably contain an axial vessel of considerable size, 

 this is true of the thicker fasciculi only. With this exception, 

 the vessels form networks, just as in other parenchymatous 

 structures, permeating the cellular bands in all directions. More- 

 over all the vessels of a sarcoma, including those of larger 

 calibre, have as good as no proper walls ; they present themselves 

 in transverse sections of hardened specimens as simple canals, 

 tunnelled and drilled through the tissue. 



The small-celled spindlc-ccll sarcoma grows by preference from 



