556 TRANSACTIONS OF SECTION RF. 
personal equation was measured at the end of each expedition. Each station 
was occupied on at least two nights. The stations referred to in the paper were 
el Daba‘a, Mersa Matruh, Baqbag, Siwa Oasis, and Khartum. The probable 
error arrived at for an adopted longitude (two nights) was discussed and found 
to be + 0°05. Values of personal equation and of adopted longitudes were given 
in the paper. Stress was laid on the utility of the method of equal altitudes in 
field work. 
Local attraction in the prime vertical has been observed by Mr. Wade at the 
two ends. of an east and west line from Helwan to Daqshur, 15 kilometres 
distant. Local attraction was first detected by mutual azimuth observations. 
The value found was confirmed by longitude determinations carried out just in 
the way described in the previous paragraph, except that an acetylene signal- 
ling lamp with occulting shutter communicating with a chronograph replaced the 
electric telegraph. To obtain further confirmation Messrs. Curry and Wade 
established a chain of seven stations along the line (produced somewhat each 
way) and determined intervals of longitude by a method of differential observa- 
tion, for details of which reference must be made to the full paper. The 
easterly local attraction at Daqshur (relative to Helwan) was found to be: 
By first method—8/''8, 
By second method—6/'9, 
By third method—7/-7, 
Intermediate stations gave for the most part intermediate values. 
Messrs. Wade and Curry made a similar set further south at Biba. As a 
result of increased experience the accuracy attained was considerable. On 
- several occasions the probable error of a longitude interval (one night) was as 
low as + 0°02. The local attraction on the Biba line was extremely small—a 
result which is in accordance with azimuth observations. 
The differential method used should prove valuable in the detailed study of 
local attraction. 
(iv) The Precision of Field Observations for Latitude. 
By B. F. HE. Keetina, M.A. 
This paper gave the results of observations made with the purpose of 
evaluating the real precision of the field latitudes of the Geodetic Survey of 
Egypt. The field programme consists of four pairs of stars, observed on at 
least three nights, the instrument used being a 10-inch Repsold theodolite. 
At an early stage in the survey it was found that the probable error on a 
single night was in an average case about one-tenth of a second, but that 
unfortunately the agreement between the nightly means was not so close as was 
to be expected from such probable errors. Further investigation was thus 
evidently necessary, and accordingly the latitude of the survey point in the 
grounds of the Helwan Observatory was observed at approximately monthly 
intervals for the space of a year, using a procedure identical with that followed 
in the field. In all eleven series of observations were made. 
The results showed very strikingly that the mean monthly latitudes were 
much less concordant than we have a right to expect from the individual 
probable errors. The monthly residuals ranged from minus nine-tenths to plus 
nine-tenths of a second, and five out of eleven were greater than four-tenths of 
a second. The probable error of a single month’s observations, as computed 
from the eleven series, was five-tenths of a second, and this appears to be about 
the proper probable error to assign to the field latitudes of the Egyptian 
Geodetic Survey. 
Attention was called to an analogous case at Kimberley in the South African 
Survey, where two series of latitudes, determined at an interval of six weeks, 
differed by nearly eight-tenths of a second. 
In the concluding part of the paper the question was discussed whether 
the above results should be followed by a change of procedure in the Egyptian 
Survey. 
