Jandaby 28, 1921] 



SCIENCE 



95 



twisting B^ around the objective mounting, 

 and noting the results. As the field is dark, 

 quite faint stars serve. 



Before use the level was carefully calibrated. 



An observation is made by setting the tele- 

 scope, pointing east, so that the image of a 

 known star passes the intersection of the 

 cross-hairs, starting a stopwatch, stopping the 

 watch by a clock, reading both ends of the 

 bubble; then pointing west and repeating. 

 This gives the instants of two passages of the 

 star across a small horizontal circle of about 

 2° or 3° radius; the mean of these is the clock 

 time of transit over the meridian. If there 

 is a change in level reading, this is allowed 

 for by the formulas for the method of equal 

 altitudes, e.g., Comstock's " Field Astronomy," 

 par. 64, equations (108) and (109). As but 

 one star is used, the correction terms depend- 

 ing on declination vanish. Of course the best 

 results are obtained with stars which pass 

 very near the zenith, they being very near the 

 prime vertical. The computations are almost 

 as simple as those with a meridian transit 

 instrument. With the arrangement used, the 

 interval between upward and downward pas- 

 sages is about 16 minutes. 



At Manila ten or twelve of the ten-day 

 stars in the American Ephemeris are bright 

 enough and culminate near enough to the 

 zenith for this apparatus. I have made a 

 good many trials, of which a large number 

 were unsatisfactory, the concrete sidewalk on 

 which the tripod stood, and on which I had to 

 move about from one sighting or reading 

 position to another, not being stable enough. 

 However, a position was found where the 

 bubble moved from this cause only an un- 

 certain fraction of a division, and the results 

 in the table were there obtained. 



The columns marked " corrections for star 

 time — time-ball time" give the corrections to 

 be added to the time-piece reading, found as 

 above, to give the standard mean time (E. 

 120°), computed from the geographical posi- 

 tion (known to 0.1 sec), and the star tables, 

 or the same interpolated from the noon-time 

 fall of the Manila Observatory time-ball. As 

 is to be expected, the latter correction is gen- 



erally found to be smaller, for the time-ball 

 has to drop a short distance for its motion 

 to be perceived. 



Taking into consideration the clumsiness of 

 the attachment, the uncertainties of stopwatch 

 readings, the instability of the platform and 

 the inexperience of the observer, the table in- 

 dicates that under better conditions the 

 method would be exact. It has the great ad- 

 vantage that highly accurate adjustments of 

 collimation axis, etc., are unimportant. It 

 can be extended by observing pairs of stars to 

 give latitude as well as time. 



WiLLARD J. Fisher 



The TJNIVERSITT of the PHHilPPINES, 



ManhjA, p. I. 



THE AMERICAN SOCIETY OF NAT- 

 URALISTS 



The thirty-eiglith annual meeting of the Ameri- 

 can Society of Naturalists was held in Ida Noyes 

 Hall and Mandel Hall, Undversity of Chicago, De- 

 cember 30 and 31, 1920. 



Ajt the business meeting the treasurer's report 

 was read, showing a balance of $514.09 in the 

 treasury. 



On recommenda/tion of the executive committee, 

 the oonstitutiou was amended by adding a sentence 

 to the end of Section 1 of Article II. This section 

 now reads: 



Section 1. Membership in this society shall be 

 limited to persons professionally engaged in some 

 branch of natural history, as, instructors in natural 

 history, officers of museums and other scientific 

 institutions, physicians, and others, who have es- 

 sentially promoted the natural history sciences by 

 original contributions of any kind. Any member 

 may present to the executive committee of the so- 

 cieity, through the secretary, names of candidates 



