As the barometer in this instrument is precisely similar to the "Long Range Barometer" invented by Mr. McNeild (and which will be found described at page 48), it may be desirable to quote the following, from Mr. Hartnup's Report:--"Mr. King constructed a small model instrument to illustrate the principle. This instrument was entrusted to my care for examination, and it was exhibited to the scientific gentlemen who visited the Observatory in 1854, during the meeting of the British Association for the Advancement of Science." =36. Syphon, with Photographic Registration.=--A continuous self-registering barometer has been constructed, in which photography is employed. Those who may wish to adopt a similar apparatus, or thoroughly to understand the arrangements and mode of observation, should consult the detailed description given in the _Greenwich Magnetical and Meteorological Observations_, 1847. As the principles are applicable to photographic registration of magnetic and electric as well as meteorologic variations in instrumental indications, it would be beside our purpose to describe fully the apparatus. The barometer is a large syphon tube; the bore of the upper and lower extremities, through which the surfaces of the mercury rise and fall, is 1-1/10 inch in diameter. The glass float in the open limb is attached to a wire, which moves a delicately-supported light lever as it alters its elevation. The fulcrum of the lever is on one side of the wire; the extremity on the other side, at four times this distance from the fulcrum, carries a vertical plate of opaque mica, having a small aperture. Through this hole the light of a gas-jet shines upon photographic paper wrapped round a cylinder placed vertically, and moved round its axis by a clock fixed with its face horizontal. The cylinder is delicately supported, and revolves in friction rollers. A bent wire on the axis is embraced by a prong on the hour hand of the time-piece; therefore the cylinder is carried round once in twelve hours. It might be arranged for a different period of rotation. As the cylinder rotates, the paper receives the action of the light, and a photographic trace is left of the movements of the barometer four times the extent of the oscillations of the float, or twice the length of the variations in the barometric column. Certain chemical processes are required in the preparation of the paper, and in developing the trace. The diagram which we give on the next page, with the explanation, taken from Drew's _Practical Meteorology_, will enable the above description to be better understood: [Illustration: Fig. 28.]