231 



the hill. Had they been bedded on gravel or clay, or the softer 

 rocks that the English Observatories are generally confined to, they 

 might have been subject to dangerously irregular movements, owing 

 to the infiltration of water in the soil. But twenty years of careful 

 observation here have not detected any effect of this sort, though they 

 have shown, in the piers of the transit instrument, the existence of 

 a small annual displacement of their tops, caused apparently by a 

 difference in the expansion from temperature of the two shafts, 

 though they were purposely cut out of the same bed in the quarry. 

 But as this displacement, even at its maximum, reaches only to 

 0*001 inch, and proceeds very regularly, its effects on the observa- 

 tions may easily be guarded against. 



The second principle above alluded to, viz., that of reversal, was 

 not introduced into the Edinburgh transit in a perfectly unexcep- 

 tionable manner. At the time of its construction it was certainly 

 thought well of. But, with the usual unhappy tendency to run to 

 extremes, men had no sooner discovered that mere weight in the 

 piers, and the telescopes resting on them, was not a guarantee for 

 their perfect stability, and that the reversingwas a necessary adjunct — 

 than they began immediately to attend almost solely to this latter fea- 

 ture, and to make the instruments so slight and delicate, as to require 

 constant reversing. Especially vicious, too, was the then plan of 

 making the metal bearings, through the intervention of which the 

 instrument rested on the pier ; for they were made so small, and so 

 weak, and of such a complicated construction, that the good qualities 

 of the masonry, such as they were, became neutralized, and very 

 much larger and more uncertain errors were introduced. 



From this source arose those various fluctuations in the position 

 of the transit instrument which I had the honour of describing to 

 the Society in 1 847. They had been first detected by my predeces- 

 sor, and were finally traced up theoretically by myself, to the un- 

 equal expansion of certain adjusting screws in the Y block. Now 

 these adjustments should never have been there ; and was precisely a 

 reason why the Y block could not be firm. They were introduced 

 with the vain idea of enabling the astronomer each day to screw up 

 the instrument to perfect truth before he began his observations. 

 But Professor Henderson knew very well, that after a screw is once 

 touched, it does not attain its true bearing for days, and sometimes 

 for weeks ; and he knew also that the quantity of any error can be 

 measured numerically much more easily than it can be corrected 



VOL. III. T 



