THE SUBMAXILLARY GLAND 239 



dog there will be a reflex flow of saliva from the mouth, even 

 though the animal does not s-mdl the food. The afferent paths 

 are now the optic nerves. But, again, this is not all, for " when 

 the dog realises that he is being played with . . . the psychical 

 secretion of saliva ceases." And yet again, it is said that the 

 thought, or memory, of savoury food may cause a secretion of 

 saliva in ourselves. 



When small pebbles are put into a dog's mouth, the animal 

 shifts them about with his tongue and then spits them out. 

 When the same substance, crushed into sand, are put in his 

 mouth, thin watery saliva is secreted, and the sand is washed out. 

 So, also, with dry biscuits. When meat is put in his mouth, 

 thick, viscid saliva is secreted. 



Thus, although the secretion of saliva is a reflex, it is one that 

 is controlled and modified (or even arrested) in a variety of ways. 

 There is no simple, invariable, mechanical response. 



How is the control effected ? 



It was thought at one time that the control was effected by 

 the change of calibre of the arterioles. The saliva comes ulti- 

 mately from the blood that circulates through the gland. Now 

 the fibres of the chorda tympani dilate the vessels and permit a 

 more abundant flow of blood through the gland. Accordingly, 

 there is a greater flow of saliva. The fibres of the sympathetic 

 contract the arterioles, and so a decreased flow of blood takes 

 place and there is a decreased flow of saliva. The inference 

 appeared to be that the secretion of saliva was a mere filtration 

 from the blood, and depended on the pressure of the latter the 

 greater the pressure, the greater the flow. But the saliva has not 

 the same composition as the liquid part of the blood, so there is 

 more than mere filtration to be accounted for. Further, when 

 the gland is poisoned with atropin, the dilating action of the 

 chorda tympani is not affected; yet when the latter nerve is 

 stimulated in such a poisoned gland, there is no increased flow of 

 saliva. On the other hand, when it is poisoned with pilocarpin 

 there is an increased flow. 



The explanation of these experiments is that there are secre- 

 tory nerve fibres in addition to those that dilate or contract the 

 arterioles. Poisoning these secretory fibres in one way or 

 another affects the secretion of saliva. We see that there are 

 nerve fibres coming into connection with the cells of the alveoli. 



It is, then, these secretory nerve fibres, acting directly on the 

 cells, that stimulate the latter to secrete the liquid. The other 

 nerve fibres which act on the bloodvessels control the supply of 



