551 
Jidda. First we find as the mean daily rates during 4 periods of from 
4 to 6 days each separated by journeys to Mecca: 
Dent D.N.7 D.N.77 D.N.80 D.N.81 DN. 84 
Bobr = 6-09) IATA (426) 3968 AB 548 003 
Ende HR eye ede 4g Tk Goes ooo le de 
March 28: 2499 “= 005920 451226 164" == gsr eas 
mie goer ea. 60° ZE HIB 015 = eN 008 
Secondly the accidental deviations have been examined, first by 
i? pe 
forming the mean value —V LAA of the differences A between 
Nn 
two subsequent daily rates, and afterwards by comparing the rates 
between Febr. 6 and March 23 themselves with their mean value 
en te 
for the whole period and deducing the mean deviation — /ZA'A'. 
n 
Both these mean deviations I and II follow here. 
Dent DENKER DEN. 77 DENT SO DIN, SP DUN, 84 
ee Wer Oot 097 O60 Ora. SOG 
eek ee LA ede 0.69 A ate Or OH 
For D.N. 81 the mean deviation I becomes + 0557, if one time- 
determination is excluded. 
The striking things in these comparisons are in particular the con- 
siderable acceleration of D.N. 7, owing to which also the mean 
deviation II is very great; and secondly the regularity of D. N. 84. 
6. Derivation of the difference of longitude Jidda—Mecea. 
v te AY | 
From the corrections and rates of our watches given in the pre- 
ceding paragraph we must now deduce ihe most probable value for 
the difference of longitude between Jidda and Mecea. Apart from 
the desirability of knowing the result yielded by each of the watches 
an immediate combination of the results of all would be impossible, 
because of the fact that on the different journeys different watches 
were taken and only Dent 2527 was used throughout. We shall 
therefore derive separately the results to which the 6 employed 
_ watches have led, and only afterwards we shall endeavour to derive 
from the whole of this material the most reliable final result. 
Whereas each group of observations at Jidda or at Mecca usually 
includes time-determinations on 4 nights, determinations on 11 nights 
at Jidda, viz. from Jan. 25 to Febr. 12 immediately precede the 
