TRANSACTIONS OF SECTION A. 541 



2. Tlie Essentials of a Machine for the Accurate Measurement of Celestial 

 Photograj)hs. By A. R. Hinks, M.A. 



It is now within the power of amateur astronomers to do work of the highest 

 value by measurinu: photographs made by the existing telescopes of the public 

 observatories in such numbers that they cannot all be measured and discussed at 

 the observatories themselves. When this is more fully realised there will be some 

 demand for a suitable measuring machine at a not extravagant price. The author 

 attempts to define the essentials of the simplest machine which will do work of 

 the highest accuracy. 



The machine shall measure one coordinate at a time on plates impressed with 

 a standard 5 mm. r(5seau. 



To ensure that the error in the measure due to the machine and the observer 

 shall not be a large part of the whole error, the machine must read to O'OOOl of a 

 reseau interval 11. 



The object glass of the microscope should project the image of a Tt-square with 

 magnification unity on to a divided glass scale in the eyepiece, to divide it into 

 one hundred parts. This scale should have the spaces numbered, not the divisions. 

 R-lines and star discs are then referred to the centre of the scale space nearest 

 each by a micrometer screw, which may be applied (1) to the plate carriage, {2) to 

 the scale, (3) to the objective. The last has not been done, but it promises the 

 advantage over the others that it brings the micrometer head at a convenient 

 distance from the eye ; and since the range of motion required is small (Oo mm. 

 is ample) the objective could be carried in the centre of an arm pivoted at one end 

 and pressed against the screw at the other, which would be simple to make. 



The objective must give a flat field over at least 5 mm. The tube carrying 

 the eyepiece and scale must have a focussing movement, preferably independent of 

 the objective, which should have a email independent range of adjustment to 

 make the R-square fit the scale and reduce errors of run. This does not disturb 

 the focus if the objective is midway between plate and scale. 



The plate carriage must move on two rectangular slides, which need not be 

 really true. It may be moved by hand, but a quick rack and pinion motion is 

 much better. Clamps are not necessary if the carriage is counterpoised. Rough 

 setting scales with adjustable pointers are necessary. The plate should be 

 brought up by springs under three studs, and an orientating screw at one corner 

 is required. 



Uniform illumination is given by a simple convex lens below the plate and a 

 concave mirror. It is most important that the observer should be shielded from 

 direct light by black curtains and screens, that he may be able to keep both eyes 

 open. 



The essentials are: (1) objective giving a flat field, and (2) divided scale in 

 the eyepiece, good optical work ; (3) micromfiter screw motion to subdivide the 

 scale spaces, the only part which wants really good mechanical work ; (4) simple 

 focussing movement ; and (5) orientating screw for the plate. 



Semi-essentials, which quickly pay for themselves in time saved and fatigue 

 avoided, are : — 



(6) The adjustment of objective independent of microscope tube by a divided 

 head ; (7) rack and pinion motion for the plate carriage ; (8) lens and concave 

 mirror illumination ; (9) light screens. 



For a discussion of most oi these points reference is made to a paper by the 

 author, ' Monthly Notices of the Royal Astronomical Society,' 1901 May. 



3. Note on the Singkep Commutator. By David P, Todd, 



At last year's meeting of the Association I described the Tripoli commutator, 

 a mechanical device which I employed at that station on May 28, 1900, for 

 operating the eclipse instruments automatically. The fortunate accident of 

 locating the instruments on the roof or terrace of the British Consulate made it 



