52 REPORT — 1876. 



that the materials chosen for compensation should, of all those not otherwise oh- 

 jectionable, he those of greatest and of least expansibility. Therefore, certainly, 

 glass or platinum ought to be one of the materials ; and the steel of the ordinary 

 astronomical mercury pendulum is a mistake. Mercury ought to be the other (its 

 cubic expansion being six times the linear expansion of zinc), unless the capillary 

 uncertainty of the mercury surface lead to irregular changes in the rate of the pen- 

 dulum. I'he weight of the pendulum ought to be of material of the greatest spe- 

 cific gravity attainable, at all events unless the whole is to be mounted in an air- 

 tight case, because one of the chief errors of the best existing pendulums is that 

 depending on the variations of barometric pressm-e. The expense of platinum puts 

 it out of the question for the weight of the pendulum, even although the use of 

 mercmy for the temperature compensation did not also give mercury for the weight. 

 Thus even though as good compensation could be got by zinc and platinum as by 

 any other means, mercury ought, ou account of its superior specific gi-avity, to be 

 preferred to lead for the weight of the pendulum. 



I have accordingly now made several pendulums (for tide-gauges) with no other 

 material in the mo\dng part than glass and mercuiy, with rounded Icnife-edges of 

 agate for the fixed support ; and I am on the point of making four more for two 

 new clocks which I am having made on the plan which forms the subject of this 

 communication. I have had no opportunity hitherto of testing the performance of 

 any of these pendulums ; but their .action seems very promising of good results, 

 and the only untoward circumstance which has hitherto appeared in connexion 

 with them has been breakages of the glass in two attempts to have one cairied 

 safely to Genoa for a tide-gauge made by Mr. White to an order for the Italian 

 Government. 



As to the acciu'acy of my new clock, it is enough to look at the pendulum vi- 

 brating with perfect steadiness, from month to month, through a range of half a 

 centimetre on each side of its middle position, with its pallets onlj'' touched dm'ing 

 ^Jg of the time by the escapement-tooth, to feel certain that, if the best ordinary 

 astronomical clock owes any of its irregidaiities to variations of range of its pen- 

 dulum, or to impulses and friction of its escapement-wheel, the new clock must, 

 when tried with an equally good pendulum, prove more regular. I hope soon to 

 have it tried vdth a better pendulum than that of any astronomical clock hitherto 

 made ; and if it then shows irregularities amounting to -fV of those of the best as- 

 tronomical clocks, the next step must be to inclose it in an air-tight case kept at 

 constant temperature, day and night, summer and winter. 



(hi Mr, Sahine's Method of Measuring small Intervals of Time. 



On Tidal Ojyerations in the Gulf of Catch hi/ the Great Trigonometrical Survey 

 of India. By Capt. A. W. Baird, B.E. 



The primary object of the operations was to determine whether secular changes 

 in the level of the land at the head of the Gulf, i. e. the " Runn of Cutch," are 

 taking place. Col. Walker, the Superintendent of the Great Trigonometrical Survey 

 of India, at first intended to restrict the observations to a few weeks duration ; but 

 he found that by extending them to a period of a little over a year, scientific re- 

 sidts of the highest value would be obtained, and also that this course would be 

 necessary in order to obtain data sufiicient to detect minute changes in the relative 

 level of land and sea. I was deputed when in England in 1871 to study the details 

 of tidal observations and harmonic analj'sis as recommended by the British Asso- 

 ciation; at the same time I tested a new self-registering tide-gauge, the per- 

 formances of which were very satisfactory. The self-registering tide-gauges were 

 then described at length, the most remarkable feature in them being the unusually 



