THE NONSPECIFIC REACTION 57 



into the fluid in the peritoneal cavity. The colloids depressed the rate ; 

 crystalloids increased the permeability when studied under these condi- 

 tions. Similar studies were made by von den Velden. 



Siegert as well as Schmidt studied particularly the effect of pro- 

 tein injections on the smaller vessels. They found that small doses 

 increased the permeability, while large doses decreased the rate. The 

 most recent studies are those of Starkenstein. These were carried 

 out by producing corneal ulcers in rabbits and then observing the 

 rate at which dyes (sodium fluorescin, etc.) would diffuse out at 

 the site of the lesion after a variety of nonspecific injections. Apart 

 from milk, albumoses and collargol he tried out a large number of 

 drugs including quinin, atophan, salicylates, adrenalin, methylene blue, 

 iodin, etc. According to his observation practically all these sub- 

 stances were followed by a diminution of the permeability of the 

 capillaries and he accounts for their antiphlogistic effect on this 

 basis. 



From the effect noted on the lymph system and the experiments 

 on the capillaries above described it seems very probable that the 

 effect of the nonspecific injections is to increase the permeability 

 of the capillaries for a short period and later to cause a definite lessen- 

 ing of the permeability. The reaction and the direction of its maxi- 

 mum effect will depend largely on the dosage employed and to some 

 degree on the agent used. 



The Lymphagogue Effect. It is well known that certain sub- 

 stances, which Heidenhain classified as lymphagogues of the first class, 

 among them peptone, egg albumin, tissue extracts, etc., cause a marked 

 increase in the lymph flow, supposed to be derived largely from the 

 liver (Starling) . This increased flow may continue a considerable time 

 following such injection. Teague and McWilliams have recently ad- 

 vanced the explanation that this lymphagogue effect is responsible for 

 the therapeutic effect of protein injections in that the antibodies of the 

 blood stream are forced into the lymph spaces and there destroy the in- 

 vading organism. Davis and Petersen investigated this effect on 

 the lymph flow, using large dogs in whom a lymph fistula was estab- 

 lished at the thoracic duct and then injecting killed colon vaccine in- 

 travenously to produce the shock effect. When necessary small doses 

 of morphin were given. If the animals were injected too soon after 

 the operation and before complete recovery had been made from the 

 anesthetic, considerable resistance to the shock was manifest and 

 the temperature reaction was delayed for several hours. 



Lymph Volume. The increase in the rate of the flow of the lymph 

 followed immediately upon the injection and in severe intoxications 

 two periods of maximum flow occurred, the first immediately after 

 the injection and persisting for from 20 to 30 minutes, the second 

 after approximately 1 hour, the latter increase being continued over 

 a longer period of time. When the intoxication was not so great the 



