734 



NATURE 



[February 3, 1921 



temperature gradient between the cistern and the 

 stem of the barometers, demand radical alterations 

 in itsdesig-n if it is to be rendered cllicient. For the 

 brief period of the voyage when the gravity baro- 

 meter could t)e favourably observed, results were 

 obtained which seemed to indicate that the method 

 was one of promise, and it may be of interest to 

 state that new apparatus of this type is in course 

 of construction ; the design has been modified 

 in the light of experience gained at sea, and of a 

 mathematical examination of the instrument by Sir 

 Arthur Schuster. One specially favourable feature 

 of the instrument is the possibility of completely 

 immersing it in a constant-temperature bath. 



Though the particular aneroid "pumped" with 

 the motion of the ship more than was hoped, and 

 had a reduction factor and zero which changed 

 slightly during the voyage, the aneroid method is 

 one which should be further examined, and certain 

 directions in which alterations in the design are 

 desirable were indicated by the experience on the 

 voyage to .Australia. With it the general varia- 

 tion of gravity with latitude over the ocean is 

 readily shown, but whether it may be trusted to 

 discriminate between such variations as mav be 

 found over deep and shallow waters must be a 

 matter for further examination. We may note, as 

 a matter of interest, that such indications as were 

 obtained with this method suggested a defect of 

 gravity over great ocean depths, along contin- 

 ental seaboards (especially when there was a 

 coastal range of mountains), and an excess of 

 gravity at island stations ; but, as we have stated, 

 a more rigorous test with improved apparatus is 

 necessary before this can be accepted. The prob- 

 lem has therefore arrived at an interesting stage; 

 Hecker's observations are in favour of nearlv full 

 compensation, whereas the slight evidence of the 

 later work, so far as it goes, suggests that com- 

 pensation is incomplete. 



Reference has already b(?en made to the instru- 

 ment constructed in 1859 by William Siemens — 

 in that year he was carried in a warship across 

 the Bay of liiscay, his real object being the deter- 

 mination of ocean depths, which he took for 

 granted would be shown by a diminished value 

 of g. Dissatisfied with his first apparatus, he 

 did not make a further attempt until 187^. when 

 he constructed an instrument which depended upon 

 balancing the pressure of a column of mercury 

 against the tension of a spring, and this he 

 tested on a cable-laying ship over a portion of 

 the voyage between the Thames and Nova Scotia. 

 The results, in spite of anomalies as regards lati- 



tude variation, which puzzled Siemens, show a 

 surprising measure of agreement between pre- 

 dicted and observed depths, which, so far as they 

 go, are in accord with the aneroid observations just 

 referred to. This must not, however, be over- 

 emphasised, since Siemens was dissatisfied with 

 this apparatus. Though not really directed at 

 the solution of the problem under discussion — 

 Siemens 's "bathometers" were graduated in 

 fathoms — these instruments are of interest in that 

 they appear to have been the first involving 

 gravity measurements to be submitted to an 

 actual test at sea. 



Since the .\ustralian meeting of the British 

 Association in 1914, further work has been 

 carried out under the auspices of an in- 

 fluential committee of that body, and certain 

 other points have received attention. From 

 a series of experiments carried out last year on 

 H.M.S. Plucky, it appears that the ship's motion 

 through the air may very appreciably affect the 

 pressure recorded by an open barometer, even 

 when carried in cabins below deck ; hence, as the 

 "lag" of this instrument is in general different 

 from that of the instrument with which it is being 

 compared, it is very undesirable to adopt baro- 

 meters of the open type for gravity determinations. 

 On board the destroyer effects as large as one 

 millibar were found to be due to the relative motion 

 of the ship and the air; no doubt a similar dis- 

 turbing influence affects the readings of a baro- 

 meter in a building about which a wind is blowing. 



Another matter which was examined on 

 H.M.S. Plucky was the Eotvos effect; going east 

 with the earth, the centripetal force is greater than 

 when steaming west ; consequently a correction 

 for motion in longitude is indicated. After elim- 

 inating windage effects, a change equivalent to 

 01 mb. was observed when the course was altered 

 from K. to W. when steaming at 22 knots. 



There are other points the investigation of 

 which is not yet complete : the best diameter 

 of capillary tubing to be used in the barometer tube 

 to damp down the effects of the ship's vertical 

 motion, the influence of the throbbing of the 

 ship's engines upon the barometer reading — there 

 is some suspicion that certain divergences between 

 gravity readings in harbour and in the open sea 

 may be accounted for by the change in capillary 

 forces due to this cause — the best form of con- 

 stant-temperature chamber for use at sea, steady 

 to i/ioo degree: these and allied questions are 

 engaging the attention of those who are contem- 

 plating a fresh attack upon the problem. 



Obituary. 



Prof. H. .A. Bi'mstead. 

 "■ I ""HE death of Prof. H. A. Bumstead. professor 

 -L of physics in Yale University, which occurred 

 with tragic suddenness on January i when he 

 was travelling from Chicago to Washington, will 

 be felt with the keenest regret by a large number 

 of men of science in this country. There are 

 NO. 2675, VOL. 106] 



few .\merican men of .science with more English 

 friends than had Prof. Bumstead, and none whose 

 friendship and companionship were more highly 

 prized. Born in 1870, he graduated at Johns 

 Hopkins in 1891. He began in 1893, as instructor 

 in physics in Sheffield Science School, that con- 

 nection with Vale which continued without inter- 



