348 



and the other when near the west prinie vertical, with but little elapsed 

 time between the periods of their being at nearly equal altitudes when 

 in these respective positions. 



When, owing to peculiar circumstances, these time stars could not 

 be observed when in the prime vertical, care was taken to observe them 

 both on the same side of it, either both when north or both when south 

 of that plane. 



This was done to secure the same eliminations of errors as are 

 above described, and also for the elimination of the effect of any small 

 error in the observed latitude, when it was to be used as a term in the 

 equation for computing the time. 



In some instances equal altitudes of the same star were observed 

 east and west, or before and after passing the meridian, for computing 

 the time. This method ensures a perfect elimination of all the errors 

 above described in obtaining the time, provided the temperature re- 

 mains nearly the same during the elapsed time between the east and 

 west observations. If, however, there be much change of tempera- 

 ture between these two periods, the effect upon the angular positions of 

 the mirrors, with respect to the plane of the circular arc, may be such 

 as to produce an appreciable error in the times of the western obser- 

 vations. One-half of this error would enter into the result for the 

 time, by chronometer, of the meridian passage of the star. There is 

 another inconvenience which sometimes attends this method of getting 

 the time, which is, that after the observations are made east, clouds 

 may arise before the time arrives for making those west, and thus cut 

 them off. There is usually an elapsed time of five or six hours within 

 which such an impediment may occur, unless the weather be settled. 



It is only, therefore, when the temperature of the outward air is pretty 

 uniform, and the weather clear and settled, that we have resorted to 

 this m thod. 



With these precautions a practical observer may, even with so light 

 and portable an apparatus as a sextant of good construction, reading 

 to ten seconds of arc by aid of the vernier, an artificial horizon of quick- 

 silver, and a good watch or chronometer, obtain a pretty close deter- 

 mination of the local time, and of the latitude of a place, by a few 

 hours' work in a sinn;le nij^ht. 



The station occupied for the observations here reported, was the 

 side walk on the north margin of Huron street, eighty (80) feet east 

 of the middle of Wolcott street. It is also four hundred and three 

 (403) feet south of the parallel, and two hundred and nine (209) feet 

 west of the meridian of the observiufr station No. 1, near the Roman 



