334 REPORT — 1851. 



strument having been so variable, no constant estimate can be given for it ; 

 in an average adjustment, however, it is believed that the probable error of 

 a measurement will not exceed 0"000015 of the whole vertical force. 



Failures in the registration have been mentioned whenever they have oc- 

 curred, and a note taken of the cause when such is known. On examination 

 of these records, I find that, in the case of the declination, there are about 

 seventy-five hours in the six months during which no registrations have been 

 obtained, owing to insufficiency of the photographic process ; in the horizon- 

 tal force there are about fifty hours. No failures whatever have taken place 

 in the photographic process during the last ten weeks of the trial. Failures 

 have occasionally occurred from causes purely accidental, such as omitting 

 to wind the clocks, not properly adjusting the sliding plate-frames, forgetting 

 to open the valve of the declination lamp, and such like. These, however, 

 must be considered rather as personal than instrumental errors. 



All the photographic registrations have been copied upon gelatine tracing- 

 paper. I am not yet prepared to give any estimate as to the accuracy with 

 which these copies are made. 



III. General Remarks on the Capabilities of the Instruments. 

 In forming an opinion as to the powers of the instruments, it is necessary 

 to take into consideration the circumstances connected with their construc- 

 tion. The declination magnetograph was the first instrument constructed 

 according to your design, and consequently cannot be expected to equal in 

 accuracy or convenience those afterwards made. From the essential portions 

 of its structure being altogether of wood, it would be too much to expect 

 from it a long-continued series of trustworthy records, where steadiness and 

 permanency of adjustment are so necessary. The large dimensions of the 

 magnet and its appendages would, on any system of observation, present great 

 difficulties whenever the more rapid magnetic changes occur. A consider- 

 able loss of light is sustained by the position of the instrument requiring the 

 daylight to be reflected, and the photographic difficulties are accordingly in- 

 creased. The want of a copper damper sufficiently powerful to eliminate 

 the mechanical oscillation of the magnet, by permitting an almost continu- 

 ous minute vibration, tends to diminish the sharpness of outline in the 

 trace. It has accordingly been found that, owing to changes taking place in 

 the framework of the apparatus, a series of more than a few weeks cannot be 

 obtained without some slight adjustments of the instrument. When the 

 magnetic changes become very rapid and extensive, it fails to aiFord all the 

 information which is desirable. Instances of failure in this respect will be 

 found in some of the larger disturbances which have occurred during the 

 trial, as, for example, in those of the 3rd and 29th of September, when, 

 from the excessive abruptness of the magnetic motions, and the want of 

 delicacy in what may be styled the registering-pencil, the interpretation of 

 the photographic records becomes very difficult and uncertain. Notwith- 

 standing the defects which I have alluded to, it is, however, certain that 

 the instrument is capable of aff"ording a large amount of information. It 

 exhibits with much exactness all the common fluctuations ; and there can 

 be no doubt that very trustworthy results, as to the mean diurnal move- 

 ments, could be obtained from it. Even disturbances of considerable 

 amount, especially those which, although of large extent, are not of an ab- 

 rupt character, are recorded with as much accuracy as seems to be desirable. 

 In short, the instrument, even in its present state, is capable of providing a 

 very great porportion of the data required for magnetical investigation. 



In the horizontal-force magnetograph, the defects of construction in the 





