ix THE SKIN AND CUTANEOUS GLANDS 519 



increase, sometimes decrease, sometimes no variation in weight. 

 These differences are readily explained by the numerous sources 

 of error inherent in the method :- 



() Imperfect sensibility of the balance used for weighing the 

 subject ; (6) difficulty of avoiding absorption by small accidental 

 lesions of the epidermis and by the mucous surfaces of the genital 

 organs ; (c) practical impossibility of estimating the amount of 

 fluid which is not absorbed but imbibed by the superficial layers 

 of the epidermis, particularly the skin of the palm and sole, which 

 is not protected by sebum ; (d) impossibility of estimating the 

 losses from the body during the bath, by respiratory gas-exchanges 

 and cutaneous transpiration, and by drying after the bath from 

 the rubbing off of the epidermis. 



More probable results, which approximate better with the facts, 

 are obtained if, instead of weighing, the method of changes in the 

 volume of the water is adopted, as shown by the displacement of 

 the level in a capillary tube, communicating with the closed 

 vessel in which a part only of the body is immersed. Fleischer 

 (1877) employed the glass cylinder of Mosso's plethysmograph 

 (which has a capillary manometer), and kept his forearm immersed 

 in the water for three hours, but found no absorption by the skin. 



In order to test whether substances soluble in water were 

 absorbed by the skin, compounds which could be chemically 

 detected in small quantities in the urine were employed. 



This chemical method is certainly the most delicate and the 

 most easily applied in the study of absorption, but integrity of the 

 epidermis in the part of the body experimented on is essential, 

 and substances which have an alterative effect on the skin must 

 be excluded. 



The majority of the experiments by this method prove defi- 

 nitely that when the epidermis is intact it lets through neither 

 water nor substances dissolved in it, provided these really have 

 no chemical action on the epidermis. Braune (1856), on steeping 

 his feet in solutions of potassium iodide, potassium iodate, iodic 

 acid, was unable subsequently to detect iodine in the urine. 

 Parisot (1863), besides baths of potassium iodide and ferrocyauide 

 solution, also tried infusions of belladonna, digitalis, and rhubarb, 

 repeating the experiment twice a day for three to eight days, 

 without finding any absorption. Hiifner (1880), after soaking his 

 feet in solution of lithium chloride, looked for lithium without 

 success in the urine by the spectroscopic method. These negative 

 results were confirmed by Winternitz (1891), who used aqueous 

 solutions of 10-15 per cent lithium chloride. Lastly, Fubini and 

 Pierini (1893) found no absorption by the skin with aqueous 

 solutions of 3 per cent potassium ferrocyanide, 2 per cent sodium 

 santonate, 5 per cent sodium salicylate, 5 per cent potassium 

 iodide, 2 per cent lithium benzoate. 



