540 



NA TURE 



[January 20, 19 10 



vertical spider-lines; one is adjustable by the 

 left drum-head of the micrometer, so as to 

 be able to set it at any convenient distance from 

 the other in order to include a single band and most 

 of the reference ring between them ; and both are 

 moved together by the other (right) measuring drum, 

 in order to be able to determine the band-width and 

 am fraction of a band which may have passed the 

 reference centre. 



The dispersion apparatus consists of a Hilger con- 

 stant-deviation prism, which enables the desired 

 spectrum ray to be isolated from all others, and that 

 alone delivered to the interference apparatus. The 

 rays are deviated exactly at right angles by this prism 

 towards the interference apparatus, the surfaces of 

 which they strike at normal incidence, after which 

 they return through the Constant-deviation prism (thus 

 securing double dispersion) to the telescope. The 

 prism is mounted on a divided circle, so that it may 

 be calibrated for the delivery of light of any desired 

 wave-length, if desired, and has numerous adjust- 

 ments. Such calibration is not essential, however, 

 as the particular image of the origin-stop in the 

 colour corresponding to the spectrum bright line of 

 cadmium or hydrogen can be adjusted visually on 

 removal of the front lenses of the Ranisden eye-piece. 



The interference apparatus consists of three circular 

 and thick glass discs, the third of which is of black 

 glass polished an absolutely true plane on its outer 

 surface, which is one of the two important surfaces 

 concerned in the production of the interfering light. 

 It is ground on the back surface, by which it is 

 attached in an adjustable manner to the right micro- 

 scope of the comparator, the movement of which it is 

 to record. The other two are larger discs of colourless 

 glass, identically similar, the two truly plane surfaces 

 of each disc not being strictly parallel, but inclined 

 at the minute angle of 35 minutes. The left surface 

 of that one nearest to the black glass disc is the 

 second surface concerned in the interference, and 

 approaches the black glass within a millimetre ; the 

 second is a duplicate one, merely introduced on the 

 right of it to correct for the slight dispersion pro- 

 duced by the 35' of inclination, the two being set 

 oppositely as regards the direction of the wedge. The 

 3j' inclination is just adequate to deflect out of the 

 field of the telescope the reflection from the other 

 (right) surface of the left colourless disc, and both 

 images from the countervailing disc are got rid of 

 by a slight tilt in the rectangular direction. All the 

 many adjustments required are provided for in the 

 mounting of the two colourless discs, on a separate 

 carrier sliding along the face of the upper V-and-plane 

 bed of the comparator. 



The apparatus as described up to this point is the 

 interferometer. 



The comparator consists of two V-and-plane beds, 

 nearly 7 feet long, of specially homogeneous 

 cast-iron, and worked truly plane with consum- 

 mate care, together with their contents; they are 

 .irranged step-wise, one on the top of the stone block, 

 and the other 75 inches below and in front. On the 

 upper one slide the two duplicate microscopes, and on 

 the lower one the standard-bar carriage and accessorv 

 fine-adjustment fittings. The carriage is given a 

 longitudinal motion, a transverse motion adequate to 

 bring either of the two bars to be compared under 

 the microscopes, as well as fine adjustments for 

 azimuth, height, and level, thus enabling the defining 

 marks on the bars to be readily focussed without 

 touching the microscopes if it is so desired. 



Each microscope is carried on a solidly constructed 

 slider on the V-and-plane bed, by which its coarse 

 adjustment for position is effected. The microscope- 

 NO. 2099, VOL. 82] 



bearing bracket is not, however, fixed directly to this 

 slider, but to a second one sliding over the first, also 

 with V-and-plane contact, and with the further con- 

 trol of the movement of a cylinder within a cylindrical 

 boring. The fine-sliding is effected by means of a 

 most carefully made screw of fifty threads to the 

 inch, on which the success of the instrument depends, 

 and which carries at its outer end a large milled 

 head for hand rotation, and a worm-wheel of 100 

 teeth gearing with an endless screw, which can either 

 be rotated by hand by means of a milled head or by 

 means of a shaft and a large wheel seen in front irf 

 the illustration. One complete rotation of the latter 

 corresponds to the movement of the microscope and 

 the black glass interference disc to an extent which 

 causes the passage of fifteen interference bands past 

 the reference centre. More than an inch of move- 

 ment of the circumference of fHe wheel is necessary 

 to effect the passage of a single'band. Two-thirds of 

 the dead-weight of the microscope and slider are 

 taken up by four spring pistons, and the movement 

 of the slider by the screw is only a push in either 

 direction against the walls of a recess in the free 

 slider, there being absolutely no strain anywhere. 

 Hence this movement of the microscope is not only 

 an excessively fine one, but also so steadv that the 

 bands pass with a precision which leaves nothing to 

 be desired, and each band may be held for any length 

 of time for counting purposes. 



Each microscope is provided with a micrometer eye- 

 piece, with spider-lines arranged as in the interfero- 

 meter. The fine adjustment is made exceptionally 

 steady and regular. Two sets of objectives are 

 provided, one pair for observing the defining lines in 

 the countersunk wells near the ends of standard bars, 

 with a magnification of 150 diameters, and without 

 penetration of the well by the objective, and the other 

 set for use with the wave-length rulings. 



The defining lines, of whatever character, are illu- 

 minated (with "critical illumination") by the brilliant 

 image of a distant Nernst lamp, with the aid in each 

 case of a little reflecting prism, a collimating lens, 

 an iris diaphragm, and a glass-plate mirror above 

 the objective, all provided with fine adjustments. This 

 avoids all heating effect on the bars, and the last 

 traces of heat rays are filtered out by a thick water- 

 jacket in front of the lamp and its beam-parallelising 

 lenses. The illumination of the wave-length rulings 

 one-forty-thousandth of an inch apart is excellent with 

 the i/i2th inch dry objectives employed, and the 

 definition truly surprising. 



The temperature of the whole comparator room is 

 maintained at the official temperature, 62° p., 

 entirely electrically, both as regards artificial heating' 

 and the thermostat, which is original. So sensitive is 

 the latter that the entrance of a person into the room 

 is immediately followed by the extinction of one of 

 the heating lamps to compensate for the extra warmth 

 introduced. 



The finest defining lines yet emploj'ed on any line- 

 measure bars are those on the platinum-iridium copy 

 of the imperial standard yard, '^'et even each of 

 these has a thickness equivalent to fifteen interfer- 

 ence bands. The defining lines on the imperial 

 yard itself are three times as coarse. Hence 

 we have now arrived at that stage in the 

 competition between defining lines and refinement 

 of measurement when the latter has far surpassed 

 the former. It was for this reason that the writer 

 took up the investigation of wave-length rulings, 

 with the idea of their possible use as defining lines 

 commensurable with the increased refinement of 

 measurement. Mr. H. J. Grayson, of Melbourne, 

 whose wonderfully fine rulings have recently been 



