and Hydnjiatle balanet. roi , 



that prefs sjently at the ends, is brought always to have the fame bearing upon the cryftal ; fo 

 that no error need be feared fron\ a fmall deviation from the right-angular- pofition of the axis 

 to the beam, fliould any fuch exift, and from its fhape and quality, it may be confidered as in- 

 flexible in ordinary experiments. At />, is a fmall adj ufting fcrew, which raifes or deprefles a 

 weight within ; and with it, in confequence, the centre of gravity in the whole beam ; by this 

 means the motion in its centre may be brought to almoft any required degree of fenfibility. 

 Should the centre of gravity be raifed above the centre of motion, the beam would turn over ; if 

 it be in that centre, the beam would ftand every where indifferently, without any. vibration; if 

 it be placed much below it, the vibration would be too quick, and its fenfibility not fufficient ; it 

 is therefore brought by the fcrew /> a very fmall quantitybelow the centre of motion, fo as to 

 defcribe one vibration in 40 or 50 feconds : the fenfibility is then fully fufficient. At each 

 end of the beam are circular boxes, « and a, through which pafs the fleel centres, from whence 

 are fufpended the fcale pans q and r : thefe centres refemble, in fome degree, thofe at »2, but 

 have their chamfered or angular edges upwards, and thereon hang the hooks /3, to which are 

 affixed the links a, and to them the three filken lines of the fcale. Each of thefe centres has 

 a motion in its refpeftive box, by means of two fmall adjufting fcrews; that in laterally, and ' 

 that in n vertically ; the former to make the two arms of the beam of an equal length, the latter 

 to bring the three points of fufpenfion of the beam and fcales into a right line. At the extre- 

 mity of the boxes are fixed two needle points, or indexes, which play againft the ivory fcale of 

 divifion at s and /. Thefe divifions, although they do not, indeed they cannot, fhew any de- 

 finite weight, are neverthelefs very ufeful in making the adjuflments, and even in weighing. 

 to the fmall fraftions of a grain, u v are two fteady plates that are raifed or depreffed by the 

 wooden nut w, to check the vibrations of the fcales q and r, and bring them fpeedily to an equi- 

 librium ; ^ z is a table, whereon the whole is placed, to raife it to an height convenient for 

 experiments*. 



To ufe with this beam I had three fets of weights made, viz. 



The firfl fet, or feries, of fifteen weights, rifing in a duplicate progreffion, from one to 16384 

 grains, viz. 



No. ■ Grains. ' 



1 = - - - - I 



2 = - - ' - - 2 



3 = - - - - 4 



* This beam is conftrufted after that made by Ramfden for the Right Honourable Sir Jbfeph Banks, Bart. 

 P.R.S. &c. and known by the name of the Royal Society's Balance. It is to be regretted that we have no 

 good account of that inftrument. The chief differences between this, and the beam defcribed in the text, are, 

 r, that certain pieces, or rings, fimilar to thofe called Jiops, in tclefcopes, are driven into the cones, and greatly 

 prevent flexure, by preferving the circular ftftion. For a tube can fcarcely bend, unlefs its cavity be made 

 10 change its figure : and 2, the planes which bear the fulcrum are fet and ground together, . and: are levelled 

 by a fpirit level placed on the face of the ftoneslthemfelves.— N. , , . ^ v '. 



