342 MATHEWS. 



dilation, and that the nerve does not cause secretion by action 

 on the gland cells ; for it is known that the drug does not 

 directly paralyze the hypothetical secretory fibers, or the gland 

 cell. To obtain the details of the drug's action, I injected into 

 the jugular vein of a medium-sized dog o. i gr. of physostigmin 

 sulphate. But although the chorda was divided, a spontaneous 

 secretion began which stimulation of the chorda considerably 

 increased. This discrepancy from Heidenhain's results is prob- 

 ably due, I believe, to the impure calabar extract he used. 



I endeavored to ascertain whether the presence of blood in 

 the capillaries was an essential condition of the post-mortem se- 

 cretion by forcing the blood out with air. After ligaturing the 

 carotid artery and placing in it a canula directed headwards I 

 rapidly cut off the head and allowed air to pass into the carotid 

 under a pressure of lOO mm. of Hg. The first experiment gave 

 a positive result. On stimulating the chorda a brief, scanty se- 

 cretion was obtained which quickly ceased. Examination of the 

 gland showed it to be practically bloodless. In two other simi- 

 lar experiments the post-mortem secretion was greatly reduced 

 in amount and ceased after i to 3 minutes, instead of lasting for 

 from 3 to 5 minutes, as normally. The glands in these experi- 

 ments still contained blood in the veins. The experiments indi- 

 cate, I believe, that the presence of blood in the capillaries is an 

 essential condition of this secretion. I regret not having been 

 able to bring my experiments to a more satisfactory con- 

 clusion, but it is to be hoped that the important bearing of this 

 post-mortem saliva upon the theory of secretion may lead to 

 its being made the subject of careful investigation. 



From the following experiments the following conclusions 

 may be drawn relative to this post-mortem secretion : 



I. After clamping the gland artery, or cutting off the head, 

 a secretion may be obtained from the submaxillary gland on stim- 

 ulating the chorda. This secretion is most abundant in the first 

 minutes, and thereafter rapidly diminishes. After four or five 

 minutes no more secretion can be obtained. The total amount 

 of saliva secreted varies from 0.3 to 1.5 cc. (Experiments 

 XVIII, XXII and LXIV.) 



