DEPARTMENT OE MERIDIAN ASTROMETRY. 



157 



of our two determinations is not liable to any error which can exert an ap- 

 preciably injurious effect upon observations of the stars. 



It was indispensable that we should actually determine whether the circles 

 had received any injury during the transfer. Accordingly we made deter- 

 mination of the graduation-correction of points 15° apart, as carefully as 

 could be effected within ten days. The result might show whether the circles 

 had received a severe shock at any time sufficient to deform them at any point. 

 The following comparison of results at Albany and San Luis, respectively, 

 indicate that the circles arrived absolutely intact : 



Determinations of Gradnatio7i -corrections. 



The mean difference of these two determinations, point by point, is 

 .046", which immediately settles the main question. There is no evidence 

 whatever of a sensible deformation, since the largest difference (assuming the 

 whole of it to be due to a real difference, and not merely to error of observa- 

 tion, as it well might be) is exactly the hundred-thousandth part of an inch. 

 The determination at San Luis, designed merely as a test, is far less elaborate 

 than the original observations at Albany to compute general tables of gradua- 

 tion corrections. Allowing for this, we may assume from the preceding 

 evidence that the probable error of determination of each of these points at 

 Albany is ± 0.02" and at San Luis ± 0.03''. It is unlikely that either proba- 

 ble error is really so small. Owing to the small number of comparisons, the 

 probable error of the differences can not be computed with accuracy, A 

 second trial of equal merit might show larger differences. But it is certain 

 that the table of graduation corrections derived in Albany are applicable 

 equally well to the readings of the instrument in San Luis, and that they are 

 probably of unusual accuracy. 



My return to Albany was by way of Europe, where conference was had 

 with various astronomers concerning fundamental observations in prospect 

 under the initiative of the Paris Congress of last April. This is of great 

 interest in connection with our own project for obtaining fundamental ob- 

 servations both at Albany and at San Luis. A part of my attention was 

 devoted to an examination of records of older observations that may prove 



