1(30 Transactions of the Society. 



VII. — On the Measurement of Grayson's New Ten-Band Plate. 

 By A. A. C. Eliot Merlin. 



(Bead December 21, 1910.) 



Mr. Grayson has recently improved his ruling machine, and has 

 kindly sent me a new ten-band plate which I have measured, with 

 the most perfect appliances, and found to be much better than his 

 earlier productions, good as these are. The new plate exactly 

 resembles the older form, the ten groups running from yoVo to 

 Toooo °f an i ncn . an d such is its accuracy that I greatly doubt 

 whether the old-fashioned methods of micrometry, still almost 

 invariably employed, could reveal the very slight inequalities 

 which exist in the spacing. No human work is, or can ever be, 

 perfect, but it is doubtful if any production of the kind has hitherto 

 approached so nearly to perfection. 



The plate has been most carefully standardized with a picked 

 semi-apochromatic objective of I.M.P. 143 on 10-inch tube and 

 1*32 N.A., used with a W.A. of 0"95 and a screen. The micro- 

 meter is the Nelson-Powell screw-traversing form, which facilitates 

 accurate setting and reduces such errors to a minimum. This was, 

 of course, supported on a separate pillar-stand, and arranged so as 

 not to actually touch the tube of the Powell No. 1. When a plate 

 consisting of varying bands has to be measured, the most suitable 

 optical combination is such as will yield the best average results 

 over the entire range of rulings. With the magnification selected, 

 the interval between the lines of a y^oo m - s P acm g was found to 

 equal about field/3, while probably field/4, or field/5, would have 

 been chosen as likely to yield the best results had such spacings 

 been alone in question. Such a power, however, would have 

 proved a serious and unnecessary disadvantage on all the other 

 nine bands, and therefore, could not be reasonably selected for the 

 purpose of standardizing the whole plate. 



The annexed table * (first set of columns) gives the results 

 obtained. Each spacing throughout the plate was spanned twice, 

 the " fixed " wire being first set by screw to the lower edge of a 

 line, and the interval measured to the lower edge of the next ; the 

 setting was then repeated from upper to upper edges of the lines, 

 and the mean of the two readings registered as the interval 



* The table herewith gives the results of Mr. Merlin's measurements in sum- 

 marized form. The complete tables of measurements are kept in the Society's 

 library, and are available for reference. 



