520 



NATURE 



[February 15, 191 2 



simultaneously along a parallel of latitude, and, if 

 possible, to combine these with regular scries of 

 ascents at places distributed as nearly as possible alonjj 

 a meridian. When this has been done a firm founda- 

 tion for a survey of the atmosphere will have been 

 laid. 



There is a certain lack of coherence about the pre- 

 sent work, so that although each individual contribu- 

 tion is excellent, the collection does not reach the 

 same standard. Something of this kind is perhaps 

 inevitable where different authors, separated from each 

 other by the Atlantic, undertake to write different 

 sections of a scientific report which are closely related 

 to each other, and require to be published without 

 undue delay. E. Gold. 



PRECISION OF LEVELLING OPERATIONS.' 



THE volume referred to below, containing the 

 account and discussion of the precise levelling 

 operations in India from 1858 to 1909, is published at 

 an opportune time. The revision of the main lines 

 of levels in this country and the establishment of 

 really permanent bench marks is, we understand, a 

 task that our Ordnance Survey intends to take up 

 at an early date. The experience gained in the 

 Indian work as recorded in this volume cannot fail 

 to be of great value. 



.As with any other physical measurement, we find 

 in the case of levelling that increased precision means 

 that problems unimportant or often unthought of 

 in earlier days rise to prominence and demand solu- 

 tion. Thus at the very outset of the subject we are 

 confronted by a question of definition ; what do we 

 mean when we say that two points are at the same 

 level? Do we mean that the distance of each point 

 from the surface which would correspond with the 

 mean sea surface, assuming the land to be removed, 

 measured along the normal is identical, or do we 

 mean that our two points lie on the same equipotential 

 surface? The former definition gives us the so-called 

 " orthometric " height, while the latter gives what has, 

 perhaps not very happily, been called the " dynamic " 

 height. 



Thus consider the case of a lake. The dynamic 

 height of every point on the water surface is evidently 

 the same, but the actual vertical distance above sea- 

 level varies from point to point, the rate of variation 

 being a maximum along a north and south line and 

 zero, if we exclude second-order distortions of the 

 spheroid, along an east and west line. 



Authorities vary as to which system is on the whole 

 the more convenient for practical use, so that the 

 Indian Survey has followed the sale plan of printing 

 both values. We may, however, venture the remark 

 that a convention which assigns different "levels" to 

 different points upon the surface of still water is 

 repugnant to a very large class of practical men, 

 namely, the engineers. The difference between the 

 heights of a station, measured on the two systems, 

 amounts to a maximum of nearly two feet in the case 

 of Bangalore, 3000 feet above the sea, a figure which 

 would obviously be largely exceeded if the levelling 

 were extended to regions of great elevation, and if 

 the mean latitude were differently selected. It is not 

 quite clear in choosing a mean latitude of 24° for 

 the zero of their dynamic heights, and thereby making 

 the system valid only for India, that the Survey 

 experts have adopted the best course. It is an argu- 

 able question, which we merely mention here without, 

 be it understood, expressing any definite opinion, 



1 " Account o( the Op»rations of the Great Trigonometrical Survey of 

 India." Vol. xix., Levelling of Precision in India (1858-1909). By Colonel 

 S. G. Birr.-ird, R.E., F.R.S. Pp. xiii + 484+xviii plates. (Dehra-Dun : 

 Office of the Trigonometrical Survey of India, 1910.) Price 10.8 rupees. 



NO. 2207, VOL. 88] 



whether, it dynamic heights are to be used at all, they 

 should not be based upon a universal datum, anil 

 therefore referred to a mean latitude of 45°. 



The discussion of the level errors is of great interest 

 and importance. The conclusion arrived at is that for 

 the Indian work the error of a circuit varies neith< . 

 directly as the length nor as the square root of the 

 length, but in accordance with an intermediate 

 formula : — 



Error in feet = \/ (ooo4)'M -h(o*ooo34)*M», 



where M is the distance in miles. 



This gives one-tenth of a foot for a line of 235 miles, 

 and one foot for about 2800 miles, a very satisfactory 

 degree of precision. 



The importance of both accurate and permanent 

 bench marks is rightly it^^i^^ed upon. Many cases 

 have been found where tHe marks have moved, and 

 obviously no deductions can be drawn as to elevations 

 or subsidences in the earth's crust unless the stability 

 of the bench marks is beyond suspicion. 



E. H. H. 



W 



DR. HENRY TAYLOR BOVEY, F.R.S. 



E announced with regret last week the death 

 of Dr. H. T. Bovey, late rector of the Imperial 

 College of Science and Technology, and formerly 

 dean of the faculty of applied science in McGill 

 University, Montreal, which occurred at his residence 

 in Eastbourne on February 2. The funeral service 

 was held at St. John's Church, Eastbourne, on 

 February 6, and his remains were interred in Ea- 

 bourne Cemetery. 



Dr. Bovey was born at Torquay in 1852, and afi' 

 being educated in a local school, entered Oueei, 

 College, Cambridge, in 1870. He graduated in 1873 

 as twelfth wrangler, and was elected a fellow of his 

 college in 1876. He entered the profession of 

 engineering, and joined the staff of the Mersey Docks 

 and Harbour Board. Whilst at Liverpool he took 

 part in founding the Liverpool Society of Civil 

 Engineers, and he had every reason to look forward 

 to a prosperous professional life in England. But an 

 accident occurred which gave his life a new ben' 

 and afforded opportunity for a brilliant career el- 

 where. Like the best type of Cambridge honou 

 man. Dr. Bovey was a keen supporter of athletii 

 Whilst taking part in a game of football, he w; 

 thrown down and had several ribs broken. He made 

 a good recovery, but one lung had been slightlv 

 injured, and he was advised to spend the next win! 

 in a dry climate, lest the wound should become 

 focus for pulmonary disease. He therefore accepted 

 from Sir William Dawson, principal of McGiH 

 University, the offer of a chair in civil engineerii^ 

 and applied mechanics, but declined to bind him^• 

 to hold this post for longer than a year. 



When Dr. Bovey arrived in Montreal in 18S1, 1 >. 

 found that his post was indeed a sinecure. Not onlv 

 was there no laboratory of any description, but his 

 chair was attached to the "Arts" faculty, and his 

 subject had to compete with literary subjects as an 

 option for a degree. At that time in McGill the prin- 

 cipal qualification for the success of an optional sub- 

 ject was constituted by its claims to be considered 

 a " soft snap," i.e. by demanding light work and 

 having easy terminal examinations. The mathe- 

 matical teaching provided by the University was quite 

 unsuited to engineering students, and Dr. Bovey's 

 efforts to have it modified met with no success. Next 

 year, therefore. Dr. Bovey resigned his chair, and was 

 about to return to England, but he was pressed by 



