1^2 PHILOSOPHICAL TRANSACTIONS. [aNNO I786. 



Being still disappointed in many instances, when, on a review of a nebula 

 whose place I had before determined, I perceived a difference of 4 or 5 minutes 

 in polar distance, I began at last entirely to new model the machinery of the 

 polar distance piece, and on Sept. 24, 1785, completed one with the following 

 capital improvements. My former piece showed a set of numbers whose value 

 differed in every situation of the telescope, and therefore required different and 

 very extensive tables to cast them up in degrees and minutes. This shows at 

 once both the degree and minute of the polar distance of every celestial object, 

 without requiring any tables to cast up numbers. In the next place, the consi- 

 derable inaccuracy arising from the unequal tension of the great ropes, and their 

 expansion or contraction by moisture or dryness, is entirely taken away ; for now 

 my index cord is contrived so as to go off from the front of the telescope itself, 

 in the direction of a tangent to the arch it describes when moving; by which 

 means this cord will even serve as an hygrometer to show the variations of the 

 ropes that suspend the telescope. If a shower of rain, for instance, should 

 shorten them so as to elevate the telescope 2, 4, or Q minutes, which has hap- 

 pened sometimes, notwithstanding they have all been well saturated with oil, the 

 index cord will immediately make the polar distance clock show this effect of the 

 rain, by pointing out an equal change of the dial. As to the variations of the 

 cord itself, they are in the first place very trifling, since it consists merely of a 

 few threads of hemp, very loosely twisted, well oiled, and always equally 

 stretched ; but especially these variations are of no consequence, as they are so 

 easily to be discovered by the check of the quadrant of altitude affixed to the 

 telescope, or the successive transits of known stars, and may either be immedi- 

 ately corrected by the adjustable hand of the polar distance dial, or be left to be 

 accounted for afterwards. 



The improvement of the right ascension has not been less attended to; and 

 the R. s. having kindly intrusted me with an excellent time-piece, I succeeded at 

 last by means of the addition of the following apparatus. Against the side of 

 the tube is fixed a vertical iron plate, and the point of suspension of the telescope 

 is disposed so as to permit this plate to be just in contact with a roller which 

 remains fixed during the time of a sweep. There is also a considerable spring 

 applied on the opposite side, in such a manner as, by always exerting a pressure 

 nearly uniform, to cause the iron plate to rub against the fixed roller as the tele- 

 scope sweeps up and down. By this means I have frequently, in very stormy 

 weather, observed many hours without finding my time materially affected, and 

 the corrections will seldom, in accurate observations, exceed a few seconds. 



To those who are accustomed to the accuracy of transit instruments in regular 

 observatories, this telescope, notwithstanding the above-mentioned improve- 

 ments, may perhaps appear far from being brought to perfection; but they should 



