ON THE KEW OBSERVATORY. 135 



III. Experiments made at the Kew Observatory in 1 843 and 1 S^^. 

 I sincerely hope that we have not wasted much of the sum granted for the 

 support of this establishment at the last meeting of the Association, in endea- 

 vouring to prosecute what we conceive to be one of its chief objects. Some 

 of the experiments (here selected from a large collection) were absolutelj'- 

 necessary (to authenticity), others yet imperfect may possibly become com- 

 plete and useful, as they may be further pursued, and some are or may be- 

 come completely useless. None of them are comprised in the many trials 

 which were made previously and more or less subservient to the construction 

 of the principal conductor and its appendages, or to the several improve- 

 ments already described of other instruments employed in the observations. 

 They may perhaps, in conjunction with the Journal, <S:c., serve at least to 

 show that sufficient precautions have not hitherto been taken for conducting 

 electro-atmospheric observations to even approximative comparability/, and 

 may possibly tend to induce far more able inquirers to favour us with whole- 

 some advice and assistance. In fact this result has already been in some 

 measure obtained in the instance of our zealous and able friend Dr. Robinson, 

 and several very eminent professors. 



1. Comparison of Voltaic Electrometers. 



Two glass pillars (called a and b), similar to F (fig. 2), were mounted, 

 with their collars, &c., upon a broad wooden shelf in the recess of the southern 

 window of the southern room D (fig. 1) ; each was provided with its warm- 

 ing-lamp, chimney, &c., and an arm projecting horizontally from the cap, 

 which arms supported the pairs of straws, &c. of the voltaic electrometers 

 to be compared (as in fig. 4), and the caps Avere placed in good conducting 

 communication by a wire. The electrometers, &c. were charged (by an elec- 

 trophorus) as highly as they could be without causing the straws to strike 

 the sides, and their divergences were not noted down until the straws had 

 somewhat collapsed. The electrometers A and C are of Volta's first or stand- 

 ard kind ; B and D of his second kind. 



Insulator a. Insulator b. 



Time. Electrometer A. Electrometer C. 



Feb.l 20° 20° 



15 14-5 



10 10 



5-25 5 



Electrometer B. Electrometer D. 



Feb.3 90 90 



80 80 



67-5 70 



47-5 50 



27-5 30 



17-5 20 



Electrometer A. Electrometer C. 



Feb.6. 13h48m 20 20 



16-5 16 



10-5 10 



This experiment (or set of experiments) suffices to show, that our ordinary 

 voltaic electrometers possess a tolerable approximation to comparability. [It 

 is difficult to estimate a much smaller quantity than 2°*5 of the electrometers 

 B and D, or half a degree of A and B.] 



2. Comparative insulating Powers of two Insulators. 

 All things remaining as in experiment I, the electrometers were charged, 

 and after time had been given them to fall a little, the caps of the insulators, 

 a and b, were contemporaneously deprived of conducting communication by 



