94 
MINUTES OF PROCEEDINGS OF 
later by Professor Chauvenet,* Professor of Mathematics, &c., at the 
Washington University, St Louis. These investigations appear to exhaust 
the subject, but very few observers will take the trouble to make the 
requisite observations to determine the errors of their sextants'at all angles. 
On the North American Boundary Commission several sextants were used by 
different observers, and with all of them the following fact appeared ; The 
latitude by southern stars was always from 40" to 80" greater than that by 
northern ones, from which it is evident that there was generally an error 
of from 20" to 40" in a result from northern or southern stars only. A 
corresponding discrepancy was also found between the times by stars east 
and west of the meridian. When the altitudes of the time stars were 
corrected by half the difference between the latitudes and the times re¬ 
computed, they invariably agreed very closely. Moreover, sextant latitudes 
were frequently compared with those obtained by the zenith telescope, 
an instrument of undoubted power, having an aperture of three inches and 
a focal length of 40 inches. The greatest discrepancy in such com¬ 
parisons was 6", but in several instances it was as small as 1". Prom these 
comparisons it became evident that with a good observer a set of sextant 
observations, such as are suggested, would give a latitude differing less than 
5" or 500 ft. on an average from a result obtained by the most powerful 
instrument. If the sun only were observed and the only correction applied 
to the observed altitude, that on account of index error a difference of 
about 30" or 3000 ft. might be expected. Stars are somewhat but not 
much more difficult to observe than the sun, and to make the observations 
two lamps are required, one for the observer, and the other for the recorder; 
the oil for these lamps should be of the best quality. None but those who 
have experienced it can know the unsatisfactoriness and loss of time and 
temper consequent upon attempting to make observations with lamps that 
burn badly. With management and a practised recorder one lamp (Bull's 
eye) may be made to answer, but it is more convenient to employ two. 
When the observer is obliged to record his own observations and to note 
the chronometer times, the following method will be found accurate and 
simple :—Either hold the chronometer in the left hand or place it in such a 
position that its beats can be distinctly heard. At the moment of contact of 
the two objects in the field of the telescope commence counting the beats of 
the chronometer, beginning at 0, which beat should be that most nearly 
coincident with the time of contact. Before counting up to 10 there will be 
plenty of time to turn the attention to the face of the chronometer, and the 
position of the seconds hand at the tenth beat should be carefully observed. 
Subtract 4 seconds from this observed time (as pocket chronometers beat 10 
times in 4 seconds) and you have the accurate chronometer time when the 
observation was made. This method of observing the time is applicable 
to almost every description of observation. Its only objection is that it 
occupies four seconds of time. 
Spherical and Practical Astronomy, Yol. II. 1863. J. B. Lippincott and Co., Philadelphia. 
