NATURAL PHILOSOPHY. 171 



my .assistant and myself each made five scries of ten observations, each on 

 alternate minutes, which being continued several days, showed that we were 

 liable to a variation of ' eye personality ' amounting to about two hundredths 

 of a second on the mean of ten observations. I also found that the differ- 

 ence already established between Mr. Twitchell and myself was confirmed in 

 in these observations. I will simply remark that the sense of touch gave 

 results almost identical with those of the eye ; and this fact being soon dis- 

 covered, the observations for personality of touch were discontinued. Thus 

 far we have presented results obtained by the eye, in seizing an almost 

 instantaneous movement, the sudden darting of a white line from behind a 

 black screen. When a comparison was instituted between the absolute and 

 observed moments of transits of the artificial stars, I found, much to my 

 surprise, that both my assistant any myself largely anticipated the true time ; 

 and that in every instance, without one exception, the same fact was noticed 

 in other persons, who Avere ignorant of what they were doing, while record- 

 ing their transits. After learning the fact of this unconscious anticipation, 

 efforts were made to cure the evil by special attention. To some extent this 

 was done, but the tendency was to an immediate relapse the moment special 

 attention was discontinued. I find (on a mean of ten observations) my own 

 anticipations amounting to the tenth of a second of time in more than one 

 instance, while Mr. Twite-hell's error is nearly as large. The variations from 

 day to day, and from observation to observation in the same set were far 

 larger than I had anticipated. This gave rise to the observation of ' emer- 

 sions ' of the artificial stars from behind a dark screen. Here I found a 

 steadiness in the results precisely equal to the performance of the eye as 

 already determined, which for my assistant and myself seems to be the 

 highest limit of attainable accuracy. The experiments of observed immer- 

 sion exhibited the fact of a strong tendency to anticipation, and a less degree 

 of steadiness in the work. I now became anxious to apply the discoveries 

 thus made in some practical manner to our star transits. For this purpose I 

 have constructed a diaphragm, consisting of eight occulting bars, four on 

 each side of a central spider's line. We observe the emersion from the first 

 bar, both immersion and emersion from the second, third, and fourth bars, 

 the transit of the central wire, the immersion and emersion of the fifth, 

 sixth, and seventh bars, and the immersion on the eighth bar; in this way 

 we make fifteen observations ; these bars are about two seconds of time in 

 width, and their intervals about four seconds at the equator. By observing 

 emersion and immersion, we hope to avoid any error arising from stars of 

 different magnitudes, as the larger stars will emerge sooner and disappear 

 later, a mean of the two observations giving us the place of an imaginary 

 wire between the two bars correctly." 



OX ROTATORY STABILITY AND ITS APPLICATIONS TO ASTRONOMI- 

 CAL OBSERVATIONS ON BOARD SHIPS. 



The following paper has been read before the Royal Society, by Professor 

 Baden Powell. The subject of rotatory motion, especially when taking place 

 under those combinations which are presented in the gyroscope, or free- 

 balanced revolver, has attracted much attention at the present day; and 

 though the primary mechanical principles bearing upon it had long .since 

 been understood, and acknowledged in theory, yet the practical results, to 

 which they might lead, had been so little considered, that when first tangibly 



