August 24, 18S;?.] 



SCIENCE. 



239 



A new method of investigating the flexure 

 corrections of a meridian circle. 



BY PROF. W. A, KOOERS OF CAMBRIDGE, MASa. 



The error due to refraction, the flexure of the 

 circle ilsolf, ami the astmnomioal flexure, the iliree 

 being fuiioliciiis in thcrasclves, are most prolific 

 errors respocling flexures of a meridian circle. 



The theory which suggested itself was arrived at 

 from the use on the telescope of a level of a differ- 

 ent construction from any the author hud ever seen. 

 He had been a disbeliever in a level, but this device 

 converted him into an advocate of the level. The 

 level tube is attached to a plate, and the plate 

 attached to the cube of the telescope. Then .set 

 the telescope at the north point, and reverse it to 

 the south, reading tlie circle north and south. It 

 would be much better were the point fixed upon a 

 ring so that it can be readily placed at any inclina- 

 tion. 



Results of tests with the almacantar, in 

 time and latitude. 



BY 8. C. CH.\XDLER, OF C.V.MBRIDGE, MASS. 



The instrument which has been named the 'al- 

 macantar ' was described and figured in a paper 

 presented to the association at its meeting in 1880. 

 In its general nature it is an equal altitude instru- 

 ment. A hollow rectangular trough containing mer- 

 cury revolves horizontally on an upright central pillar. 

 The trough contains a float which is perfectly free to 

 obtain equilibrium, while it is constrained to revolve 

 with the trough. The float carries a telescope which 

 turns on a horizontal axis, and can be clamped at any 

 desired altitude. When this instrument is revolved 

 on its vertical axis, any given point in the field of 

 view describes a horizontal small circle, or .ilmacantar, 

 in the heavens. The transits of stars over a series 

 of horizontal lines will thus afford m«ans of deter- 

 mining the altitude of the instrument, the error of 

 the clock, the latitude or the declinations of stars, 

 by a proper distribution of the observations in 

 azimuth. 



A higher degree of accuracy is attainable by this 

 instrument than by k transit or a zenith telescope of 

 same size. The author's comparison of results is as 

 follows: The probable error of a single star in deter- 

 mining the clock error is only ±0.0.V or ±O.OIj'. 

 With a transit instrument of the same size, the quan- 

 tity is not less than ±0.08'. With the almacantar 

 the probable error in determining the latitude of a 

 single star is ±0.."».")", including the error of the star's 

 place. . This is about equal to the probable error of a 

 pair of stars by Talcott's method, with the larger 

 telescopes of the United-States coast-survey. 



The instrument was a small one, — Ij inches aper- 

 ture and 2.") inches focus. It was constructed for ex- 

 periment only, in a provisional way, at a cost of ?iI50. 

 There are obvious defects in design and construction : 

 when these are remedied, the error can be much re- 

 duced. 



A series of observations with this instrument are 

 given by the author, for the latitude of a pier about 



80 feet north of the Harvard-college observatory. 

 The value obtained by averaging these is 0.7" less 

 than given by Professor I'eirce in his discussion of 

 the prime vortical transit observations taken by the 

 Messrs. Boiul, and adopted as the standard value of 

 the latitude of the ob.servatory. The author con- 

 cludes that Professor Peirce's value is too large by 

 fully three-quarters of a seconil. Uy way of proof 

 the author gives a series of observations on the five 

 stars useil by Professor Peirce. These are compared 

 with those of Auwers and Boss, and the correction of 

 the hitherto accepted value of the latitude now indi- 

 cated by the almacantar is thereby confirmed. 



The clock errors of two nights selected at random, 

 as given by the almacantar, were exhibited by the 

 author. The results both in time and latitude would 

 be considered satisf.ictory with an ordinary instru- 

 ment of two or three times the size. The almacan- 

 tar can be made much larger than the one under 

 triiil, certainly of five or six inches aperture, with cor- 

 responding incre.ise of precision along with greater 

 optical power. Its mechanical construction is simple, 

 and^reduces the sources of error. Thus in the older 

 instruments there are involved: 1°. The accurate 

 construction of parts, as of pivots, level, graduated 

 circles. 2°. Fixity of mounting, to avoid a shifting 

 of the instrumental plane. ;J°. Rigidity of the in- 

 strument itself, to secure constancy of collimation 

 and flexure. In the almacantar only the last condi- 

 tion has to be satisfied, and it is by far the easiest of 

 the three to be attained mechanically. 



The author regards the principle of flotation 

 adopted as being as delicate an indication of the di- 

 rection of gravity as is obtained by the spirit-level. 



The alnuaeantar gives promise of a new instrument- 

 al resource in the higher practical astronomy. It is 

 competent to deal with the must delicate problems. 

 It will evade some of the minute .sources of error 

 that still cling to meridian instruments. Especially, 

 it furnishes a method for obviating diflicullies, 

 hitherto regarded as almost insuperable, connected 

 with flexure and refraction, in observations with the 

 meridian circle. 



Internal contacts in transits of the inferior 

 planets. 



BY J. B. EASTM.VN, of WASHIXOTOX, D.C. 



TiTE author began by reviewing the different values 

 obtained in observing transits of Venus, and by com- 

 jiutations thereon since 17(!1. Eventually it became 

 certain that the differences of these values depended 

 chiefly upon the computer's interpretation of the 

 observer's record. The phenomenon known as the 

 • black drop ' began to be considered as an element 

 in the calculation. Stone regarded it a4 a necessary 

 phenomenon. lie gave an explanation of its origin, 

 and stated that the moment when a <l.irk ligament 

 anpears to connect the apparent limbs of the sun and 

 Venus is the time of real internal contact. The 

 second phase, when the limbs of Venus and the sun 

 appear in contact, Stone says, is 'the apparent in- 

 ternal contact.' 



