122 CARNEGIE INSTITUTION 



the external temperature. Therefore, if any transit circle is to be 

 sent to the southern hemisphere, and is to be used before or afterward 

 in the northern hemisphere for the purpose of comparing results, it 

 seould be provided, for both series of observations, with a modern 

 steel observatory having double or triple walls, means of conveying 

 the convection currents away from the observing opening, and of 

 separating the whole observatory into two halves, so that the instru- 

 ment may be used as nearly as possible in the open air. 



There is a still more important condition which should also be 

 fulfilled, viz., provision of means to avoid personal equations de- 

 pending on magnitude or upon the velocity of the star's motion 

 (i. e., the star's declination). It seems to me that the only system 

 of observing in R. A. which permits this possibility is the Repsold- 

 Struve method, in which a wire is made to travel across the field at 

 nearly the same velocity as the star. The eye piece travels with the 

 wire, so that, if the mechanical conditions are properly realized, 

 the observer, having bisected the star disc with the wire, should 

 view the disc so bisected apparently as if at rest, and be able by 

 simple means to correct any errors of this bisection which he may 

 notice during transit and which may be due to errors of the clock- 

 work, the driving screw, or the original pointing. The drumhead 

 of the screw which causes the slide of the moving wire to travel is 

 provided with contacts which, as the drum rotates, make electric 

 contact with the chronograph circuit, and so record the instants when 

 the pointing on the star would correspond with the particular read- 

 ings of the micrometer head. 



This method, so far as I am aware, is the only one not liable to 

 personality depending on magnitude or declination, and this, 

 although we have not yet absolute proof, we believe also to be free 

 from personality in observations of the sun and moon. 



The necessity for provision of reliable azimuth marks and of a 

 clock not liable to diurnal variation of rate is too well known to 

 require further reference. 



A separate memorandum dealing with some details of the above- 

 mentioned methods will be forwarded. 



As to a site, I think it would be difficult to find in the southern 

 hemisphere a better one for this purpose than the neighborhood of 

 Bloemfontein. I venture to think that in connection with this plan 

 one or two of the northern observatories should be provided with a 

 better form of observatory or covering for the transit circle. 



