8o6 



NATURE 



[August 26, 1920 



ings and partly to provide funds sufficient to maintain 

 and work them. 



There are clearly only two sources from which 

 the very large sums required by the universities can 

 be obtained, and those are (i) Treasury grants and 

 (2) private benefactions. 



You have pointed out that the proposed Treasury 

 grant of 1,500,000^. included in the Estimates for 

 1921-22 is quite inadequate, and it is obvious that 

 this must be the case. It is, therefore, to be hoped 

 that careful inquiry into the needs of the universities 

 by the Treasury will result in this sum being sub- 

 stantially augmented. With regard to private bene- 

 faction, I think we may look forward with confidence 

 to very liberal response in the near future from 

 generous individuals, and more particularly from 

 wealthy firms interested in the progress of science and 

 education. The action of Messrs. Brunner, Mond, and 

 Co. in setting aside 100,000?. as a contribution to the 

 universities is an example which will certainly he 

 followed by other firms who owe much of their success 

 to the work of chemists, engineers, and others sent to 

 them from the universities. 



If it were to become a recognised practice for firms 

 who can afford to do so to set aside yearly some com- 

 paratively small sum as a contribution to the uni- 

 versities, the combined effort would go far to solve the 

 difficulties in which we find ourselves at the present 

 time. W. H. Perkin. 



The University Museum, Oxford, August 22. 



Use of Sumner Lines in Navigation. 



May I venture to point out, in the interests of 

 navigational science, that although the article by 

 Capt. Tizard in Nature of July i under the above 

 title is an admirably clear and concise account of the 

 application of Sumner lines in navigation at the date 

 given in his examples, it is scarcely descriptive of the 

 best practice of to-day? 



Of the two methods of drawing the lines described 

 by Capt. Tizard, the first, or •"original Sumner 

 method," is now merely of academic interest, and is 

 seldom practised outside schools and examination- 

 rooms. Its defects are, first, that each sight has to 

 be worked out twice (once for each of the two assumed 

 latitudes), and, secondly, that it is inapplicable to 

 sights taken near the meridian. It may also be 

 remarked that unless the two assumed latitudes are 

 very close together, the true circle of position may 

 differ considerably from the straight line joining the 

 two points found on the Mercator chart. 



yhe second method described by Capt. Tizard, 

 usually known as the "chronometer method," is still 

 used to some extent. It avoids the double working 

 out of each sight, but gives good results only for 

 observations taken on large bearings ; it is inapplicable 

 to sights taken near the meridian. 



For observations near the meridian what is called 

 the "ex-meridian method " may be used to draw the 

 position-lines. In this method the longitude is 

 assumed, and the sight is worked out as a latitude 

 observation ; the position-line is then drawn, at right 

 angles to the bearing, through the point where the 

 meridian of the assumed longitude is cut by the 

 parallel of the observed latitude. This method gives 

 good results for sights near the meridian, but fails 

 on large bearings. 



A combination of the last two methods is some- 

 times employed, the sights near the meridian being 

 worked out by the "ex-meridian method," and those 

 on large bearings by the "chronometer method." 

 This combined procedure has been advocated by 

 several writers, especially by Capt. Blackburne, who 



NO. 2652, VOL. 105] 



undertook the immense labour of computing tables 

 specially adapted for the purpose. The main objec- 

 tion to it is that the procedure is not uniform for all 

 sights. 



A much better method of drawing the Sumner lines 

 than any of the above, and one which seems destined 

 to replace all others, being now in extensive use by 

 navigators of all nations and recognised as the 

 standard method in the Royal Navy, is known as the 

 Marcq Saint-Hilaire method, or the "new naviga- 

 tion." It consists in assuming a dead-reckoning 

 position in both latitude and longitude, and then 

 tinding how much the observed zenith-distance of a 

 heavenly body differs from that calculated on the 

 assumption that the dead-reckoning position was cor- 

 rect. The difference between the observed and cal- 

 culated zenith-distances is laid off from the assumed 

 position in a direction to or from the observed object 

 (according as the observed zenith-distance is less or 

 greater than the calculated one), and the position-line 

 is then drawn through the point so found at right 

 angles to the bearing. The great advantage of this 

 method is that it is perfectly general ; it gives equally 

 good results whatever the bearing of the object 

 sighted. 



rhough called the "new navigation," the Marcq 

 Saint-Hilaire method of drawing the Sumner lines 

 is by no means a recent invention, having been used 

 in the French Navy for more than forty years. Its 

 advantages were advocated so long ago as 1888 bv 

 that indefatigable worker for the advancement of 

 navigation, the Rev. William Hall, R.N., and have 

 since been frequently pointed out by other Eng- 

 lish writers on navigation. Its superiority over all 

 other methods for drawing the Sumner lines or 

 position-lines being indubitable, there is a little 

 difficulty in understanding its comparative neglect by 

 British navigators up to recent times. One reason, 

 no doubt, is conservatism ; the British seaman 

 usually prefers to use time-honoured methods with 

 which he is familiar rather than to adopt new-fangled 

 notions, and fears to risk his ship bv the possibility 

 of making a mistake in a process with which he has 

 not been made acquainted during his early training. 

 Another reason which operated powerfully until 

 within the last twenty years or so was the absence 

 of any tables for facilitating the calculation of alti- 

 tudes comparable in scope with the tables of Davis 

 and Burdwood, which so greatly helped in the rapid 

 reduction of sights by the "chronometer method." 

 This last difficulty was removed by the publication of 

 the excellent "Altitude Tables " of my namesake, the 

 Rev. F. Ball, M..^., of the Royal Navy, and at the 

 present time it is just as simple a matter to work 

 out sights by the "new navigation," with the aid of 

 these tables, as it was to work the old "chronometer 

 method " with the help of Burdwood and Davis. 



Until within the last decade it was seldom worth 

 while to attempt to fix a . ship's position at sea 

 within a mile or two, because so long as the longi- 

 tude, whether found by Sumner lines or by any other 

 method, was dependent entirely on the Greenwich 

 time as found bv the transport of chronometers over 

 long distances, it was usually impossible to be sure 

 of the longitude within that amount, no matter how 

 accurately star observations were made. This diffi- 

 culty affected the hydrographic surveyor as well as the 

 navigator; and, indeed, it provides the explanation 

 why so many charted longitudes — down the Red Sea, 

 for instance — are in error "by a mile or more. But 

 nowadays, when wireless time-signals enable the error 

 of a ship's chronometer to be found daily with an 

 accuracy of a few tenths of a second anywhere on the 

 seas, there is no reason why the longitude should 



