358 REPORT — 1851. 



vations recorded in the Journal between September 10 and 30, is about 

 four. These were generally made at about 9 and 11 a.m., and 1, 3 and 5 p.m. 

 The daily mean number between October 1 and March 28 is nearly six, 

 usually made at about 7 and 10 a.m., and 1, 4, 7 and 10 p.m., at about one 

 hour after sunrise and at about sunset. 



The mean number of daily compound observations recorded is, from Sep- 

 tember 10 to 30, about three, and from October 1 to March 28 about five; 

 the highest monthly mean is seven (in October). 



Some of the circumstances which have operated to prevent the chosen 

 number of eight daily observations having been uniformly completed, were, — 

 absence of flame in the lantern occasioned by high wind, &c. ; great 

 and irregular disturbance of the pendulums of the electrometers occasioned 

 by dust, &c. injuring the insulation, violent wind suggesting the caution of 

 securing the conductor from damage by means of stays (of cord) attached 

 to it ; violent wind and rain ; the necessary absence of the observer ; and 

 the intervention of Sundays (but even on these days three observations have 

 been almost constantly made at about 7, 9 and 10 a.m.). 



Five instances only are recorded of absence of signs, but the gold-leaf 

 electroscope was not used. These were — on January 9*^ 20*^ 50', on March 

 5^ l^ 0', on March 5"^ 2>'^ 57', on March 10'^ 4'^ 4', and on March 19"* 3^ 57', . 

 and were marked cases of disturbance. In two or three of these instance* 

 the circumstances of weather were precisely those in which the atmosphere 

 itself has been commonly believed to be totally uninsulating, viz. those of 

 great and long-continued humidity. On March 10 the readings were high 

 and negative in several observations before that at 4'' 4', and positive in seve- 

 ral afterwards ; a state of transition might therefore have existed at the ob- 

 servation at 4'' 4' *. 



Some of the above-mentioned circumstances have operated much more 

 frequently in unfitting a compound observation for employment in the fre- 

 quency deductions (entered in column I.), than in preventing an observation 

 entirely ; and some even of those entries were obtained under circumstances 

 of weather, conditions of the apparatus, &c., which render them somewhat 

 objectionable in discussion. 



This (so-called) Frequency course terminated, necessarily, at the end of 

 March (in order that the Magnetic course might commence (on April 1)); 

 but between April 14 and May 14 about forty single electrical observations, 

 accompanied by others for pressure, temperature, &c., as above, were re- 

 corded at rather irregular intervals ; most frequently at about 10 a.m., 4 p.m. 

 and 10 p.m. 



III. EXPERIMENTS, &c. 



The principal operations, &c., of which 1 set down a brief summary under 

 this head, scarcely commenced before September 1850. A classified state- 

 ment may perhaps be found more convenient than an enumeration of thera 

 in the exact order of their dates would have been. 



Horizontal-force Magnetograph. 

 On August 3 1,1 850, this instrument was rendered more sensitive than it had 



the " frequency paper," and employed in deducing the value, in inverse proportion, of fre- 

 quency set down in column I. {vide p. 354, ante). 



* Rain attended the four last-meutioued observations. It may possibly have fallen from 

 a neutral part of a cloud or stratum, electrified by induction, and may not have acquired a 

 sensible charge in its descent. 



