GLANDULAR CAllCINOMATA. 197 



increase or diminution ; we notice that they finally become dis- 

 solved and disappear after dozens of colloid layers have already 

 accumulated ; and a spontaneous conviction arises in our minds, 

 that the major part of this colloid matter has been secreted at the 

 junction of the connective tissue with the epithelium, without the 

 fully-formed epithelial cells taking any active part worth men- 

 tioning in the process. The hypothesis of a direct transudation 

 from the blood cannot be entertained for a moment, the endos- 

 motic equivalent of colloid matter being almost nil. On the 

 other hand I am inclined to adopt the view put forth by 

 Doiitrelepont (^Laiige7ibech'' s Archiv, sent in for publication in 

 April 1870), who believes the colloid substance to be a product 

 of the metamorphosis of the material which serves for the con- 

 struction of epithelial cells, an albuminous compound of some 

 kind, which would, in other forms of cancer, be employed in 

 increasing the number of the cells. We cannot but think in this 

 connexion of Arnold's theory of the origin of epithelial cells 

 from amorphous matter ; the accumulation of colloid, supposing 

 Arnold's view to be correct, admitting of being naturally ex- 

 plained as an accumulation and subsequent transformation of this 

 amorphous plasma. The concentric lamination of the colloid 

 matter points to a certain periodicity in the rate of its secretion, 

 while the fatty and granular debris marking out the limits of 

 each layer, will have to be regarded as a bye-product of its 

 manufacture. 



§ 163. It is certain that vessels ramify upon the septa and 

 trabeculae of the stroma in colloid cancer ; but these vessels are 

 neither numerous nor large, and contribute but little, even 

 when fully injected, to colour the mass of the tumour, so that 

 colloid cancer always gives one the impression of being very 

 scantily supplied with blood. As regards the course of the 

 disease, and the dangers to which the patient is liable, this cir- 

 cumstance is important, inasmuch as bleeding from the surface 

 of a colloid cancer when broken-up and ulcerating is neither 

 frequent nor abundant. 



As regards malignity, colloid cancer occupies a somewhat 

 peculiar position. It has an extraordinary power of infection 

 per contlguum. The infiltration of the connective tissue which 

 goes on at its circumference often assumes proportions which are 

 actually colossal. On the other hand, metastases to more remote 



