646 evening discourses. 



were drunk thirty minutes before collection of urine, so that samples could be 

 obtained every seven minutes or so. The samples were pumped out by means 

 of the Gardner and Buckmaster gas-pump, in which there are no taps and leakage 

 of air is practically nil. 



1 rite the results of two experiments : — 



I.- Breathed air at 3 atmospheres. After 15 minutes emptied bladder. 

 Sample I., collected 7 minutes later at 3 atmospheres. Decompressed to If atmo- 

 sphere, in 3 minutes. Sample II., collected 6 minutes later at If atmosphere. 

 Decompressed to 1 atmosphere in 3 minutes. Sample III., collected 3 minutes 

 later at 1 atmosphere. 



Sample I. 

 Sample II. 

 Sample III. , . 



II. At 3 atmospheres for 44 minutes. Emptied bladder and breathed oxygen 

 for 9 minutes, then took Sample I. Decompressed to If atmosphere in 

 2 minutes. Took Sample II. 4 minutes later. Decompressed to 1 atmosphere in 

 1£ minute. Took Sample III. 5 minutes later. 



Nitrogen 

 Sample I. 2-091 



Sample II. . 0-8835 



Sample III. . . . 0-5751 



The results show that the urine is supersaturated with nitrogen after decompres- 

 sion in the first case, and undersaturated after breathing oxygen in the second 

 case. 



The ideal method, then, for safe decompression from high pressure is (1) 

 oxygen breathing for five minutes and rapid decompression to 2 atmospheres, 

 (2) pause during which oxygen is breathed and exercise taken, (3) rapid decom- 

 pression to 1 atmosphere while oxygen breathing and exercise are continued. 



The period of decompression can be notably shortened by such means, how 

 far further experiment will show. We want to know in particular how the fat 

 of the spinal cord is desaturated under these conditions. The 'quick' parts 

 are evidently put right in a few minutes. Further experiments on fat pigs 

 should give the required information. 



Recompression. 



Recompression is the one method of cure for the illness. Pol and Watelle 

 (1854) recorded the benefit of this. Men with ' bends ' went back under pressure. 

 A. Smith suggested the use of a recompression chamber at Brooklyn. E. \Y. 

 Moir instituted it at the Hudson Tunnel. All caisson works are now provided 

 with such. Men at the East River Tunnel works have truly been raised from 

 the dead by its means. 



In the frog's web experiment I have observed the bubbles shrink up on 

 recompression. Experiments on animals show that recompression must be 

 applied at once in dangerous cases, before vital parts are killed by the inter- 

 ference with the circulation. ' Bends' may be relieved by compression long after 

 they have come on. 



Recompression relieved 90 per cent, of the cases at the East River Tunnel, 

 and all but 05 per cent, were partly relieved by its means. Oxygen breathing 

 can be employed with advantage in the medical lock. Decompression from the 

 lock must be slow, for some of the bubbles having run together to form larger 

 ones only shrink up on recompression, and do not quickly go into solution. 

 These expand again on decompression. J. F. Twort and I have observed this 

 happening, and measured the bubbles under the microscope. 



For deep-diving work a recompression chamber should always be at hand. I 

 have contrived a double-chambered diving-bell, one chamber open to the sea, the 

 other closed save for a manhole communicating with the first. The divers 

 after completing their work enter the inner chamber and close the manhole. 

 The bell is raised on deck, and the men decompressed by the stage, method. 







