Lat. N.Long. E.
°′″°′″
Boema, according to Hassanein Bey24138232440
“ “ Rohlfs[6]243138231240
Difference1830120

Hassanein Bey thus places Boema some 40 km. south-southeast of the position assigned to it by Rohlfs from Stecker’s observations. The remarkable thing about this large discrepancy is that it is chiefly in the latitude, which was directly observed by Stecker at Boema itself, and by Hassanein Bey at Taj, only 2 km. away from Boema. I have not been able to find any details of Stecker’s observations, except that they were carried out with a “prismatic circle.” But I have subjected the original records of Hassanein Bey’s observations for time and latitude at Taj to very careful examination, and there is overwhelming evidence that his latitude for that place cannot be in error by more than 1′. He observed the altitude of Polaris at Taj on no fewer than six different nights, with a watch whose error on local time was accurately known from sun and star observations carried out on the same dates. From the internal evidence of the observations themselves, the watch-error at the time of sighting Polaris cannot on any occasion have been uncertain by more than two seconds, which would not, of course, sensibly affect the latitude found; the recorded magnetic bearing of the star sighted, as well as its rate of apparent motion, proves that the star used for latitude in each case was really Polaris; and the greatest difference of any single observed latitude from the mean of the six nights’ observations was only 15″, the average departure of a single observation from the mean being 12″. Hassanein Bey’s latitude of 24° 13′ 47″ for Taj can therefore be unhesitatingly accepted as correct within 1′; and since there is not room for an error of even this amount in his estimation of the short distance of Boema from Taj, it becomes absolutely certain that Rohlfs’s latitude for Boema is over a quarter of a degree too high.

It is curious to notice that in the case of Buseima ([p. 332]) the discrepancy of 13′ 31″ between Rohlfs’s latitude and that estimated from Hassanein Bey’s recent work is of the same order and of the same sign as that found at Boema; and that a negative correction equal in amount to the sun’s semidiameter would in each case bring the results of the two observers into fairly close agreement. The explanation hence suggests itself that Stecker may have determined his latitudes by observation of the upper limb of the sun at noon, and both at Buseima and Boema failed to correct the measured altitude for the semidiameter, thus making both latitudes 16′ too great. A mistake of this kind, as every scientific traveler knows, is very easily made in hurried reductions in the field; and at the time when Stecker carried out his observations and calculations in Kufra, both he and his brave leader were in imminent peril of losing their lives at the hands of treacherous Badawi.

A similar explanation may account to a large extent for the discrepancies of longitude at the two places. Thus, according to Hassanein Bey’s work, Rohlfs’s longitude for his camp at Buseima is 9′ too far east, and that for his camp at Boema 12′ too far west. We have only to assume that Stecker observed the sun’s lower limb in the morning at Buseima, and the upper limb in the afternoon at Boema, for finding the local times, and in each case omitted to correct the observed altitude for semidiameter, to account pretty completely for both discrepancies of longitude.

The puzzling thing about the above explanation of the errors of Rohlfs’s map is that Rohlfs actually traveled the distance between Buseima and Boema, and estimated it at 120 km.,[7] whereas Hassanein Bey’s positions would indicate that the true distance is 40 km. greater. But as Rohlfs’s statement of the distance was obviously written after Stecker had determined the positions of the places astronomically, it is probable that he obtained his 120 km. by calculation from the astronomical positions, rejecting any rough estimate he may have made of the distance from his times of marching. Both Hassanein Bey and Mrs. Forbes had considered that the true distance was much greater than 120 km. when they traversed it in 1921; but as on that occasion no observations for position were taken, it remained uncertain whether either Buseima or Boema had been wrongly placed on Rohlfs’s map. It is now practically certain that both were wrong.

As regards the level of Kufra, it is satisfactory to notice a very close agreement of Hassanein Bey’s observations with those of Rohlfs. Hassanein Bey’s barometer-readings at Ezeila, to the south of Jof, give 389 meters for the level of that place, and he estimates that Boema lies some ten meters higher; this would give about four hundred meters for the altitude of Boema, a figure identical with that found by Rohlfs. Taj, which has been built on the cliff north of Jof, since Rohlfs’s day, is found to be 475 meters above sea-level from a series of aneroid-readings extending over a fortnight. The outlying villages of Kufra which lie north of Taj are lower than Taj itself, but substantially higher than the southern villages of Kufra; Awadel is 434 meters above sea, and Hawari and Hawawiri are about the same altitude.

There is also fairly close agreement as to the extent of Kufra from north to south. Rohlfs’s map makes the extent in altitude between Hawawiri and Tollab to be about 35 km., while Hassanein Bey makes it 30 km. But when we come to the distance over which the villages extend from east to west, there is a much graver discrepancy. Rohlfs makes the extent from east to west (Buma to Tollab) about 40 km., while according to the latest estimations by Hassanein Bey the true extent is only about 21 km. As Rohlfs appears to have located many of the villages on his map merely from Arab statements, and not from careful personal estimation as did Hassanein Bey, we need have no hesitation in accepting the latter’s relative positions as by far the more nearly correct, and hence concluding that Rohlfs’s map gives an east and west extension which is nearly double the truth.

The error in east and west extension (so far as concerns the placing of the villages, not the extent of the vegetation) is even greater on the map prepared by me and published by Mrs. Forbes in 1921.[8] This is due to the distance between Jof and Tollab having been much overestimated on the former journey; it was given to me as about 42 km., while according to Hassanein Bey’s latest estimation it is only 20 km.

A matter which will at once strike the eye of any one who compares Hassanein Bey’s latest map of the Kufra villages with that published by Mrs. Forbes is that in the later map a place called Ezeila is shown south of Jof, while in the earlier map (which was prepared from Hassanein Bey’s own data and rough sketches) Ezeila is shown north of Hawawiri. The explanation is that there are two Ezeilas. The name “Ezeila” is applied locally to any outlying well (usually with a clump of palms) which forms the last watering place for caravans leaving the oasis. Thus the northern Ezeila is the last well for a traveler leaving Kufra to go northeastward toward Jaghbub, while the southern Ezeila is the last well in Kufra for any one going south toward Wadai.

From the southern Ezeila of Kufra to Arkenu is 266 km. in a nearly southeasterly direction, and there is no water or grazing on the way. From Arkenu to Ouenat is a further 42 km. in a slightly more southerly direction.