96 



DISPLACEMENT INTERFEROMETRY APPLIED TO 



water. Since the total air-pressure is above 1,000 cm. of water, the accuracy 

 of reading is easily within 5Xio~ 5 . But for the secondary disturbances the 

 apparatus would thus be very satisfactory for the purpose, particularly as it 

 admits of observation in an agitated environment. 



The two divers selected for these experiments were B' and F' here to be 

 distinguished as B" and F" (fig. 112). The same figure contains the temper- 

 ature changes (t inci easing downward) and the graph p of the corresponding 

 barometric pressures in centimeters of mercury. 



The case of F" with a relatively wide tube below (0.5 cm. diameter) as to 

 absolute air-loss per day d(g/m)/dt, curiously enough, does not differ much 

 from the cases A' and C (wide-mouthed and sheathed) and is actually much 

 greater than the corresponding rate of F' (examined by exhaustion). The 

 mean value of the relative rate per day, r = 0.0015. 



The water-head excess here was but 20 or 30 cm., so that at high barometer 

 or low temperature, the diver sank permanently (s in fig. 112). Under these 

 circumstances there is always a peak in the curve, owing to increased effective 

 air-loss. This is the solutional effect of high pressure (gas being absorbed 

 by solution) and the largest rate of F" compared with F' may also be referred 

 to the same cause; i, e., the greater mean water -head under which the diver 

 is stored. Finally, the diver remained at the bottom persistently, and 

 further systematic observations were temporarily abandoned. On June 15, 

 however, another observation was made, now under a necessarily reduced 

 head, by the exhaustion methcd. The value of g/m is 91,000, showing that 

 for this diver also the rate of air-loss when the diver is left quite without inter- 

 ference (other than temperature) is as large as during the period of daily ob- 

 servations. The peculiar stationary behavior of the diver as to air-content at 

 the beginning of the above experiments (C, F', A), must therefore be attrib- 

 uted to the composition of the initial charge of air. 



The case of B" with its relatively narrow mouth (tube 0.2 cm. in diameter) 

 shows the smallest rate hitherto obtained. In fact, at the outset the air- 

 losses were negligible ; but this is attributable either to the composition of the 

 imprisoned air or to the persistent fall of temperature. The mean relative 

 rate is r = 0.0003 P er day, only about one-tenth that in most of the preceeding 

 cases. The curve is again quite irregular 

 and peaks appear when the diver is sunk 

 (s), particularly when this occurs at low 

 temperature, as is liable to be the case. 

 The present graph contains the most com- 

 plete series of examinations. I therefore 

 applied the correction for changes of tem- 

 perature of the water-head (76), this being IQtAprtt) $QJty 10 



77 



the only one of significance. The results so obtained, together with the 

 temperature graph, are shown in figure 115 in a doubled scale, g/m being, 

 as before, in ten thousands. It appears that the irregularities of the curve 

 have rather been accentuated. The solutional effect for the sunk diver (s), 



