DEDIFFERENTIATION OF TUMOURS 321 



of the lymphatic or blood-vascular endothelium. Without, I 

 trust, taking anything from the interest and value of his forth- 

 coming article upon this subject, I would here note that Dr. 

 P. G. Woolley, Fellow in Pathology at McGill University, has 

 just completed a most elaborate and minute study of a tumour, 

 originating in the zona fascicularis of the suprarenal, in which 

 a similar transition from adenomatoid to purely sarcomatous 

 structure is to be followed without the possibility of doubt. 1 



Here, then, are tumours which, showing in the least aberrant 

 regions indications of origin from a lining membrane or lepidic 

 tissue, are apt to take on the appearance and structures more 

 characteristic of " pulp " tumours. 



Now this difference in behaviour between the Epilepidomata 

 and Hypolepidomata on the one hand, and sundry of the Meso- 

 theliomata and Endotheliomata (Mesolepidomata) on the other, 

 is but consonant with embryological observations and the broadest 

 biological principles. One great principle which we see con- 

 stantly in evidence is that those structures and properties which 

 are of oldest acquirement are those which are last to be lost ; it is 

 the most recent acquirement which tends to be the earliest to dis- 

 appear. We see this exemplified continually in connexion with 

 blastomas in general. The more rapid the growth the more the 

 cells of a tumour depart from their normal mature environment 

 the more do we observe that those features of the tumour cells 

 which are specific for one or other tissue tend to disappear. 

 In the most rapidly growing and most aberrant tumours the 

 individual cells afford us little or no clue to the tissue of origin. 

 It is the general arrangement of the cells that aids us in making 

 our diagnosis, and even the general arrangement is not so much 

 that peculiar to the fully grown tissue as that common to con- 

 nective tissues in general, or to glandular or lepidic tissues in 

 general.' We recognize a reversion to an earlier, simpler, or, 

 as we express it, embryonic type. As I have pointed out else- 

 where, a distinction must be recognized between functional 

 and proliferative activity with loss of those features which 

 are directly associated with the performance of function. 



I would now suggest that we carry the working of this principle 

 a little farther back. The first lining membranes to be differ- 



1 [Trans. Assoc. Amer. Phys. xvii., 1902, 627, and Virch. Arch. 172, 

 1903, 301.] 



Y 



