The above table fails, however, to bring out fully the actual variations of the watch. So long as the other five watches remained in order, Hassanein Bey made frequent comparisons with his principal watch, and between March 21 and 23 there is strong evidence that the principal watch made an abnormal gain of about fifty seconds. A similar abnormal gain of twenty-nine seconds by the principal watch is evidenced in the twenty-four hours between comparisons of March 24 and 25. Both these abnormalities occurred between Jalo and El Harrash, at an early stage of the journey, while all the watches appeared to be behaving tolerably well, and it is quite possible that other abnormalities occurred at later stages, when, owing to some or all of the other watches having stopped or broken down, no satisfactory control by comparison was possible.

Of the other five watches carried, one was an English half-chronometer, similar to the principal watch but of smaller size; three were high-class Swiss lever (“Peerless”) watches with very tight-fitting cases; and the fifth was a small Swiss lever-watch with luminous dial, carried on the wrist for noting the times of marching. The small half-chronometer stopped on April 3, after going for over four months; and though it was restarted, its rate changed considerably after the stoppage. The three “Peerless” watches, though they failed to continue going to the end of the journey, showed by no means a bad record. One was found stopping and unreliable on May 6, after going for over five months; the two others continued to go for over a month longer; and so far as can be judged from the comparisons made on the route, their variations of rate were about of the same order as those of the half-chronometers. The wrist-watch, from the manner in which it was carried, was of course liable to much greater variations of rate, and was occasionally reset by the principal watch; but it kept going till nearly the end of the journey.

As regards dust-resisting power, which must always be one of the principal aims in selection of watches for desert exploration, there appears to be nothing to choose between good English half-chronometers and the highest class of Swiss watches, the cases of the latter being remarkably close-fitting. The most probable cause either of stoppage or of abnormal changes of rate appears to be the sudden shocks which may arise either to a watch carried on the person when jumping on or off a camel, or to one carried in the baggage by sudden movements of the camel. The most likely explanation of the abnormal gains of the principal watch for short periods on the two occasions above noted would seem to be that, owing to a jerk in mounting or dismounting, two contiguous coils of the hair-spring may have been made to touch each other for a short time, with consequent temporary shortening of the period of vibration of the balance-wheel. It is noteworthy that the watch which remained going throughout the journey was the largest of all those taken, and its greater resisting-power may have been to some extent due to its size, permitting of greater strength in its component parts.

3. Astronomical Determinations of Latitude

Observations for latitude by altitudes of the pole-star were carried out on thirty-five nights at nineteen principal camps, using the same three-inch theodolite as was employed for the time-observations. Three readings of altitude were made on each face, using each of the three horizontal stadia-wires in turn, the corresponding times being noted on a half-chronometer watch whose error on local time was accurately known from sun or star-observations carried out just before the latitude-observation. Particular care was taken with the level-adjustment, and the air-pressure and temperature at the time of the observation were recorded.

The following table gives the results of the observations:

Astronomical Latitudes

°′″
Sollum4nightsLat.3135 9N.
Siwa1““291241“
Jaghbub5““294426“
Camp near Jalo1““291156“
Jalo (El Erg)1““29 233“
Buttafal1““285426“
El Harrash1““252629“
Taj6““241347“
Arkenu2““221232“
Ouenat1““215229“
Erdi1““183539“
Agah1““175238“
Enebah1““172124“
Bao1““162824“
Furawia2““152151“
Um Buru2““15 357“
Kuttum1““141215“
El Fasher2““1338 3“
El Obeid1““131051“

Of six of the above places (Sollum, Siwa, Jaghbub, Kuttum, El Fasher, and El Obeid) the latitudes are accurately known from the Egyptian and Sudan official surveys, and the agreement in these cases is very satisfactory, though a very close comparison is not generally possible owing to uncertainty as to Hassanein Bey’s precise observation-spot. At Jaghbub, Hassanein Bey records that his observation-spot was 200 meters S.S.W. of the dome of the mosque. Applying the corresponding difference of latitude (− 6″) to my own determination of the latitude of the dome in 1917 (29° 44′ 41″) we obtain 29° 44′ 35″, showing a difference of only 9″ from Hassanein Bey’s observed latitude.

A further test of the degree of precision of the latitude-observations can be made by comparison of the latitudes found for the same camp by observations taken on different nights. The following gives the average deviation of a single observed latitude from the mean at all the camps where two or more observations for latitude were made: