THE BLOOD. 105 



ney, in that which supervenes upon scarlatina, in hydropsies 

 arising from insufficient and improper diet, as well as in 

 those which occasionally follow suddenly suppressed per- 

 spiration, it has been observed that the urine is always 

 albuminous, and the writers just mentioned have ascertained 

 the existence of a fact which stands in close relation with the 

 albuminous excess in the urine in the instances enumerated, 

 viz. that the blood is itself in these cases, on the contrary, 

 deficient in albumen. The knowledge of these two facts, 

 therefore, and the observation that their occurrence is so 

 generally associated with dropsical effusion, has led MM. 

 Andral and Gavarret to entertain the opinion that a close 

 connection exists between the deficiency of albumen in the 

 blood, and the forms of dropsy alluded to, and which is pro- 

 bably that of cause and effect. 



In speaking of non- organic dropsies it has, until recently, 

 been conceived sufficient to say that they depended as a 

 cause upon impoverishment of the blood ; but this expression 

 we now know to be vague, inasmuch as it does not convey 

 any exact notion of the real changes which the blood may 

 have undergone: we have seen that the blood may be rich in 

 its red globules or in its fibrin, and also that it may be poor 

 in these elements in almost every proportion and degree. 

 Let us see whether a deficiency of either of the two con- 

 stituents just named predisposes to dropsies. In anaemic 

 states, and in chlorosis especially, we have an impoverished 

 state of the blood, and in these we know that it is the red 

 corpuscles which are deficient; and yet daily experience shows 

 us that dropsy in such states, and particularly in chlorosis, 

 even in its most severe forms, is a very rare termination. 

 In febrile disorders again, we have, for the most part, an 

 impoverished condition of the blood, arising from the de- 

 pression in the scale of the fibrin, and yet of these we do not 

 find dropsical effusion to be by any means a frequent result. 



It is not excess of water in the blood which gives rise to 

 dropsy, for, if that were the case, then would it frequently 

 occur in the disorder to which we have" referred, chlorosis, in 



