XXXVI 



obtain, at a moderate cost, barometers and thermometers of more 

 accurate construction and trustworthy character than those usually 

 sold, directed Mr. "Welsh to undertake a series of experiments for 

 that object. The results of his labours were most satisfactory, and 

 are fully described in a paper printed in the Reports of the Asso- 

 ciation for 1853. Accordingly, at the present time, standard instru- 

 ments are supplied to scientific investigators direct from the Obser- 

 vatory, and barometers and thermometers which have been compared 

 with the standards at Kew, and each accompanied with its special 

 table of corrections, are not only supplied to the Government De- 

 partments of the Admiralty and the Board of Trade, but can now be 

 obtained from any instrument-maker at greatly reduced prices. 



In the summer of 1852 Mr. Welsh made four ascents in a balloon, 

 for scientific objects. A detailed account of these ascents and of the 

 experiments he performed, with a description of the various instru- 

 ments employed, and a statement of the general results obtained, 

 was communicated to the Royal Society in 1853, and published in 

 the ' Philosophical Transactions ' for that year. 



Sir John Herschel, in his article " Meteorology" in the 8th edi- 

 tion of the ' Encyclopeedia Britannica,' makes the following remarks 

 on Mr. Welsh's performance: "All the observations were con- 

 ducted with scrupulous precision, and the reductions very scienti- 

 fically made ; . . . these four ascents leaving nothing to desire in point 

 of instrumental appliances and scientific precision in their use." 



In 1854, Mr. Welsh, at the request of the Kew Committee, under- 

 took a series of experimental investigations on the action of the 

 mercury in marine barometers, known under the term of " pump- 

 ing." For this purpose, he went, in company with Mr. Adie, to 

 Leith and" back in a steamer, and subsequently to the Channel 

 Islands. The result of his observations led him to the conclusion 

 that the tube of a marine barometer should, in order to reduce the 

 pumpings, be contracted, so that the mercury will take about twenty 

 minutes to fall from the top of the tube to the height indicating the 

 true pressure ; and that by this means the probable error from the 

 cause indicated would not exceed 0*01 of an inch. The account of 

 these experiments was published in the Kew Report for 1853. 



In 1855, in consequence of its having been represented to the 

 Committee that Her Majesty's Government were anxious that mag- 



