CHAP, i.] TISSUES AND MECHANISMS OF DIGESTION. 551 



osmotic current passing from the weaker solution of the salt, 

 namely the blood, to the stronger solution of the salt, namely the 

 intestinal contents. But the difference between these effects of a 

 dose of magnesium sulphate and those of a corresponding dose of 

 sodium chloride are much greater than can be accounted for "by 

 the diffusion phenomena, by the differing osmotic equivalents of 

 the two substances ; and the more the matter is studied the more 

 reason have we to believe that the flow of water produced by 

 the former is to a large extent the result of suddenly increased 

 secretory activity. So also the fact that the contents of the small 

 intestine throughout its length retain the same amount of water 

 relatively to the solids, that is to say maintain the same or nearly 

 the same fluidity, whereas in the large intestine the water relatively 

 diminishes until at last the faeces become firm and even dry, cannot 

 be wholly explained without calling into our aid variations in active 

 secretion as distinguished from mere physical diffusion. And in 

 the case of a purgative such as croton oil producing a watery 

 stool, when only a minimal, we might almost say an infinitesimal 

 amount of its own substance can at any one time be present in 

 the intestinal walls, the result is obviously due to active secretion. 



If, however, we are thus driven to the conclusion that the 

 passage from the blood into the intestine is a manifestation of 

 secretory activity in which epithelium cells play a part, gradually 

 becoming little by little more intelligible to us, why should we 

 not admit that the passage from the intestine to the blood, which 

 as we have seen does not accord in its phenomena with known 

 processes of ordinary diffusion, is also brought about by the 

 activity of cells, is in fact a kind of inverted secretion, and 

 hence like ordinary secretion presents problems which cannot be 

 solved by any off-hand references to known physical processes ? 

 Indeed this is the conclusion towards which observation and 

 experiment seem to be steadily leading us. Were the alveolus of 

 a salivary gland habitually filled with a fluid of mixed and varied 

 nature like the contents of the alimentary canal, we should 

 probably in our study of the gland find ourselves compelled to 

 speak of a double current as existing in the gland, of a current 

 from the cells to the lumen of the alveolus, and of a current from 

 the lumen to the cells. And all along the intestine both the 

 columnar and cubical cells, which everywhere bear the marks of 

 being " active " cells, may perhaps be regarded as engaged in a like 

 double function. Over the villi the receptive function, in the 

 glands of Lieberkiihn the ejective function is predominant; but as 

 we have suggested, 265, in the glands reception probably is not 

 wholly absent, and we may imagine that in the villi some 

 amount of ejection (quite apart from the action of the goblet 

 cells) may take place. 



If this view be accepted, if we admit that the entrance of 

 digested food does not take place by ordinary diffusion, the question 



