STELLAE PHOTOGRAPHY. 185 



lighter than the German form in common use, since the tube is not eccentric, and 

 therefore no counterpoise is required. It does not have to be reversed when a star 

 crosses the meridian, and in fact any star may be followed uninterruptedly from rising 

 to setting. The advantage over the English form of mounting is, that polar stars can 

 be observed without difficulty. The objection to its general introduction is, that 

 the end of the polar axis would interfere with observations made near the pole. A 

 declination cii'cle is attached to the tube and divided into single degrees. A closer 

 division is not necessary, since a small error in setting would only affect the part of 

 the plate on which a star would impress itself Notches are cut in the edges of the 

 circle for each degree, every tenth notch being made deeper than the others. A 

 catch at the end of a steel spring is attached to the fork, and enables the telescope 

 to be moved quickly and with precision either 1° or 10° at a time, as may be desired- 

 The edges of the catch are ground at a greater angle than that of friction, and thus 

 permit the telescope to be moved when sufficient pressure is applied. The right 

 ascension circle consists of an iron wheel 67 cm. in diameter, in the edges of which 

 720 teeth are cut with great care. A screw turns in these teeth, forming a worm 

 and wheel. One tooth corresponds to half a degree, or to two minutes of time in 

 right ascension. An index serves to set the telescope within a minute or less, and 

 this is close enough for the present purpose. The screw is driven by clockwork 

 controlled by a Bond spring-governor, the weights being connected by the Huy- 

 ghenian arrangement, so that they may be rewound without stopping the clock- 

 work. It is difficult to make a pendulum keep accurate time in a location where 

 the variations in temperature are great. Moreover, to photograph the fainter stars 

 it is necessary that the clockwork should run for a long time with a high degree 

 of accuracy. These conditions are fulfilled by controlling the Bond spring-governor 

 electrically by a clock mounted in the clock-room of the Observatory. If the polar 

 axis is parallel to the axis of the earth, the telescope can thus be made to follow a 

 star for an indefinite time, as closely as the clock can be made to run. The error 

 from this source is therefore entirely inappreciable. The differential refraction of 

 the air would be a more serious source of error than this. The electrical contacts 

 occurred in the clock every two seconds, and that corresponding to the fifty-eighth 

 second was omitted. A spring-governor will run more .satisfactorily when the 

 contact occurs every second, or oftener ; but with the present arrangement the error 

 could seldom exceed a fraction of a second. The fifty-eighth second was omitted by 

 a secondary connection in the clock occurring once a minute, and not, as is sometimes 

 done, by the omission of one tooth in a contact wheel. The removal of the second- 



