cxviii THE BEGINNINGS OF LIFE. 



themselves in individual cases, yet there is a general re- 

 semblance which suffices to maintain the distinctive nature 

 of each affection. In this broad sense they are undoubtedly 

 entided to rank as 'specific' diseases. But though they 

 may be presumed to be associated with definite changes in 

 the blood, we have no right to infer that such changes of 

 state can only be induced in one way. Many well-known 

 chemical changes are capable of being brought about by 

 more than one agency. And just as there is the best 

 reason for believing that cancer or tubercle may be initiated 

 de novo by the operation of irritants upon the tissues of 

 certain individuals, and that such growths may subsequently 

 be multiplied within the body by the contact -influence 

 exerted by some of their disseminated particles; so may 

 we suppose, not only that specific substances (contagia) 

 may be capable of initiating specific changes in the blood, 

 but that certain combinations of circumstances may, by their 

 action upon the human body, entail similar definite changes 

 and states of blood. Having to do with a perverted nutritive 

 activity and mode of growth in a limited area of tissue, 

 cancer or tubercle may make their appearance ; whilst, 

 having an altered nutritive activity and set of changes 

 occurring in the blood, this all-pervading tissue may lapse 

 into the successive states peculiar to one or other of the 

 specific diseases, and so give rise to the symptoms by which 

 they are characterized. This is by no means a forced 

 analogy. Can cancer or tubercle arise in the individual 

 without any pre-existing 'hereditary taint'? Can the states 

 of blood peculiar to the several specific diseases arise de 

 novo, or independently of contagion.? These are questions 

 whose import is really similar ^ 



^ This double mode of causation is perfectly familiar to the chemist. 

 Particular chemical changes may occur under the influence of certain 

 general determining conditions, which at other times (in the absence of 



