Considerations on the Problem of Latitude Variation. 23 



takes then the following form, taking the origin at the highest 

 point of the level, — 



where the quantity A depends on the quality and movement of the 

 level. Under the same conditions as before, in the case of our 

 sensitive level, we have 



^ n 



This equation shows that another perturbing term interferes in 

 this case. Thus it is to be concluded that the indication of the 

 level bubble has an error greater than the resistance 6^, when the 

 level itself is in motion. 



Now I showed in the foregoing part of this essay that there 

 are reasons to believe in an unsteadiness of the telescope and pier, 

 not to speak of the seismic movement of the ground. From the 

 above investigations, we see that the level cannot indicate instan- 

 taneously the varying position of the telescope in presence of these 

 disturbances. So it is manifest that the spirit level which we now 

 use is not suited to fulfilling our requirements with the degree of 

 accuracy demanded in modern astronomical measurements. We 

 are faced with the necessity of using some other means in order to 

 realize the present expectations of practical astronomy. 



3. Application to the Talcott-Horrebow observations and 

 deduction of effect on the variation of latitude. 



In the first section of this essay I discussed certain systematic 

 motions of the levels and zenith telescope, which can be looked 

 upon as due, firstly, to the disturbance from the observer, and 

 secondly, to a terrestrial cause. In the second section I investiga- 

 ted the failure of the function of the level which can l)e taken as a 

 common defect of the level. Conversely, I can now conjecture 

 how the telescope and level behave under such circumstances. 

 When the observer approaches the telescope, the stand would first 

 suffer a thermal disturbance, and cause the said efïect, owing to the 

 unsymmetry of meteorological conditions. To this the level bubble 



