SEA-nSlIERIES LABORATORV. 139 



these contain blood corpuscles, but the number of the 

 latter is relatively few. 



Everywhere in sections of the papillae, and in the 

 tissues directly underneath, are capillary tubes ramifying 

 to an extraordinary extent, branching and anastomosing 

 very freely. They are most numerous directly beneath 

 the surface, where they form an irregular layer, but they 

 are present also in the depths : they are represented in 

 fig. 5 by the lines of dots. Some of them are shown in 

 fig. 4, and are represented as filled with short fibrils. I 

 thought at first that the contents of these vessels were 

 granular, but examination under high power lenses shows 

 the fibrillar nature of this material. In unstained 

 sections this substance is coloured light brown in mass, 

 but where a few of the fibrils can be seen at the cut edges 

 of the preparation, they appear to be nearly colourless. 

 They present a glistening appearance when seen by 

 reflected light. They do not stain with methyl-blue- 

 eosin, Mallory's combination, Ehrlich's haematoxylin 

 and eosin stain, iron haematoxylin, or eosin alone. They 

 do not Gram-stain, nor do they take stain from carbol- 

 gentian. I thought at first that they might be bacterial 

 in nature, or perhaps the hyphae of a fungus, but the 

 negative staining reaction makes these interpretations 

 impossible. They do not react in any way to dilute 

 acetic acid, and cannot be crystals of lime. They are not 

 lipoid for they do not dissolve out under treatment with 

 xylol. They are some kind of inclusion in the tissues of 

 the growth, contained either in capillary blood-vessels or 

 in channels of their own. 



The deeper parts of the growth vary in minute 

 structure. In many places what is seen is essentially 

 the condition already described in the case of the fibro- 

 sarcomatous tumour from the cod ; that is, there is a 



