446 BLOOD AND LYMPH. 



for instance, that in the limbs, under normal conditions, the flow is 

 extremely scanty, while from the liver and the intestinal area it is 

 relatively abundant. In fact, the lymph of the thoracic duct may 

 be considered as being derived almost entirely from the latter two 

 regions. Moreover, the lymph from the liver is characterized by 

 a greater percentage of proteins. To account for these differences 

 Starling suggests the plausible explanation of a variation in permea- 

 bility in the capillary walls. The capillaries seem to have a similar 

 structure all over the body so far as this is revealed to us by the 

 microscope, but the fact that the lymph-flow varies so much in 

 quantity and composition indicates that the similarity is only 

 superficial, and that in different organs the capillary walls may 

 have different internal structures, and therefore different permea- 

 bilities. This factor is evidently one of great importance. The 

 idea that the permeability of the capillaries may vary under dif- 

 ferent conditions has long been used in pathology to explain the 

 production of that excess of lymph which gives rise to the condition 

 of dropsy or edema. The theories and experiments made in con- 

 nection with this pathological condition have, in fact, a direct bearing 

 upon the theories of lymph formation.* Under normal conditions 

 the lymph is drained off as it is formed, while under pathological 

 conditions it may accumulate in the tissues owing either to an 

 excessive formation of lymph or to some interruption in its circu- 

 lation. From the foregoing considerations it is evident that changes 

 in capillary pressure, however produced, may alter the flow of 

 lymph from the blood-vessels to the tissues, by increasing or 

 decreasing, as the case may be, the amount of filtration ; changes in 

 the composition of the blood, such as follow periods of digestion, 

 will cause diffusion and osmotic streams tending to equalize the 

 composition of blood and lymph; and changes in the tissues them- 

 selves following upon physiological or pathological activity will 

 also disturb the equilibrium of composition, and, therefore, set up 

 diffusion and osmotic currents. In this way a continual interchange 

 is taking place by means of which the nutrition of the tissues is 

 effected, each according to its needs. The details of this interchange 

 must of necessity be very complex when we consider the possibilities 

 of local effects in different parts of the body. The total effects of 

 general changes, such as may be produced experimentally, are 

 simpler, and, as we have seen, are explained satisfactorily by the 

 physical and chemical factors enumerated. 



* Consult Meltzer, "Edema" ("Harrington Lectures"), "American 

 Medicine," 8, Nos. 1, 2, 4, and 5, 1904. 



