340 



NA TURE 



[February i i, 1S92 



enabled to follow the erratic variations of the former in- 

 struments through all the changes of pressure to which 

 they were subjected in the course of his travels. Armed 

 with this experience, on his return to London, he under- 

 took a series of systematic experiments, not only with 

 the instruments that he had used on the Andes, but also 

 with a large number furnished for the purpose chiefly by 

 Messrs. Hicks and Casella. These were subjected to 

 conditions which reproduced as nearly as possible those 

 experienced on the journey, and their behaviour was 

 noted under all the varying circumstances. The results 

 of these experiments were that all aneroids, when brought 

 under a low pressure, continue falling for four or five 

 weeks, and in some cases longer ; that the amount of the 

 total fall varies greatly with different instruments ; and 

 that, in general, two-thirds of the fall takes place in the 

 first week. Mr. Whymper says : " I have seen the index 

 error of an aneroid grow to as much as four inches; in 

 several instances there have been alterations of more than 

 an inch, and in numerous instances there have been altera- 

 tions from scarcely appreciable errors to + or - errors of 

 two or four tenths of an inch." On the other hand, aneroids 

 that have been kept for some weeks at a low pressure, 

 when restored to the full pressure of the atmosphere, 

 take many weeks to regain their condition of equilibrium ; 

 and when they have attained this final condition, their 

 readings are sometimes higher, sometimes lower, than 

 their original values at the same pressure. The greater 

 part of the recovery takes place in the first week, and a 

 considerable part in the course of the first day. 



Notwithstanding this sluggishness of action, Mr. 

 Whymper finds that the aneroid may be usefully em- 

 ployed for measuring differences of altitude when all the 

 readings are taken within a short interval of time ; the 

 shorter the better, so that the data compared are only 

 the first effects of the changes of pressure ; and this 

 equally whether the base station is at a high or low 

 level. He gives an instance of this in his determination 

 of the depth of the great ravine of Guallabamba, north of 

 Quito, at the top of which his mercurial barometer read 

 21-692 inches, and the two aneroids that he carried with 

 him gave readings respectively 0*552 and 0752 inch 

 too low. But when the three instruments were next 

 read, two and a half hours later, at the bottom of the 

 ravine, the mercurial barometer indicated a rise of 2'237 

 inches, and both the aneroids a rise of 2'26o inch, 

 involving an error of only i per cent. In this instance, 

 the difference of elevation as shown by the aneroids was 

 vitiated by a very small error, but the absolute heights 

 above sea-level as obtained from their readings would 

 involve errors of more than 600 and 800 feet respectively. 



That this was no accidental result was subsequently 

 confirmed by an experiment on twenty-two aneroids (all 

 having large but very varied errors). After these had 

 been kept a week at a pressure of 21*692 inches (of the 

 manometer), they were gradually restored in the course 

 of two and a half hours to a pressure of 23929 inches ; 

 and, with a single exception, the rise of the aneroid read- 

 ings ranged between 2- 130 and 2*360 inches, the mean of 

 the whole being 2*218 inches. 



As the general result of his experience, Mr. Whymper 

 concludes that all aneroids lose on the mercurial baro- 

 meter when subjected to diminished pressure, and that 

 NO. I 163, VOL. 45] 



the loss is the greater the greater the reduction of pres- 

 sure ; that when diminished pressure is maintained con- 

 tinuously, the loss commonly continues to augment 

 during several weeks, but in a constantly diminishing 

 ratio ; that when pressure commences to be restored, the 

 aneroid endeavours to recover the previous loss, and 

 some gain more than they have lost ; but the recovery is 

 gradual, and usually extends over a greater length of 

 time than the period during which the diminished pres- 

 sure has been experienced ; finally, that the index errors 

 of aneroids are never constant, so that apparently no 

 process of verification can be trusted to yield corrections 

 for permanent application, even though time be made a 

 factor of the correction formula. 



It was no part of Mr. Whymper's purpose to go deeper 

 into the matter, and to ascertain wherein lay the source of 

 the irregular action of his instruments. But it is evident 

 that this must be known before we can look for any 

 important improvement in the construction of the aneroid. 

 In all probability it lies in the varying elasticity of the 

 thin corrugated disk that forms the cover of the exhausted 

 chamber, the alternate rings of which are thrown into a 

 state of strain and stress in the process of exhaustion, and 

 which strain and stress are varied with every change of 

 the external pressure. Perhaps some clue to the cause 

 may be found in the results of Mr. Herbert Tomlinson's 

 experiments on the elasticity of metallic wires after de- 

 formation by tension or torsion,^ since he found that, after 

 such " treatment, the metal takes a considerable time to 

 recover its normal elasticity. It is, indeed, by no means 

 certain that such changes of pressure as were dealt with 

 by Mr. Whymper are sufficient to produce deformation, 

 but the aneroid affords a very delicate measure of any 

 change of elasticity in the corrugated disk, and there is 

 so much resemblance in the results of Mr. Whymper's 

 and Mr. Tomlinson's experiments as to make it at least 

 not unlikely that there is a community of cause. 



Meanwhile, travellers must bear in mind that unless 

 the aneroid can be frequently verified by comparison 

 with a mercurial barometer, its indications can be trusted 

 only for such small differences of elevation as can be 

 measured within an interval of a very few hours, A 

 rough verification can, indeed, be obtained with the 

 boiling-point thermometer, as is recommended by the 

 authors of " Hints to Travellers," and this will at least 

 enable them to avoid large and accumulated errors. 

 A fair idea of the degree of accuracy that may be 

 expected of this latter instrument in practice, is afforded 

 by Dr. Scully's simultaneous observations of the mer- 

 curial barometer and the boiling-point thermometer in 

 his journey over the Karakoram from Leh to Yarkand, 

 which will be found in vol. i. of the " Indian Meteoro- 

 logical Memoirs." H. F. B. 



WALLER'S HUMAN PHYSIOLOGY. 

 An Introduction to Human Physiology. By Augustus D. 

 Waller, M.D. (London : Longmans, Green, and Co., 

 1891.) 



IN these days, when the cult of the examination fetish 

 is in the ascendant, and we are rapidly approaching 

 the condition of the unchanging students of Confucius, 



' Phil. Trans.' i88^. 



