332 PROCEEDINGS OF THE AMERICAN ACADEMY 



The writer is using at the present time in connection with T'Gl a 

 spirit thermometer made by Mr. J. S. Iluddleston of Boston. The 

 lenirth of each desfree Centigrade is about 2.5 centimeters, and the 

 length of the column of (colored) alcohol is about thirty-nine inches. 

 The complete determination of the errors of this thermometer will 

 need to be deferred till the coming winter ; but according to the 

 present indications it is an instrument of extraordinary precision. 

 The greatest deviation thus far observed from the corrected readings 

 of rCl is about 0°.06. 



Description of Microscopes. 

 Values of One Revolution of Micrometer Screws. 



The measuring microscope used in connection with Comparator 

 No. 1 has a tube fourteen inches in length. The micrometer was made 

 by Powell and Leland. 



Nearly all of the observations have been made with a Tolles four- 

 system inch. The illumination is hivariably obtained by the use of a 

 prism inserted between the two front lenses, a device known as Tolles's 

 opaque illuminator. The principle of this illuminntor is so often stated 

 incorrectly, that it is well to restate it here. The focus of the rays of 

 light which pass through the prism is a little outside of the focus of 

 the objective itself. The image under the objective, therefore, is witliui 

 a cone of diffused light, the axis of the cone being in the line of 

 collimation of the objective. This condition, however, requires a some- 

 what careful adjustment of the prism when it is set by the maker. 



The illumination of polished metal surfaces is simply perfect. Sky 

 illumination gives better results than artificial illumination. The dis- 

 tance of the objective from the opening through which the light passes 



