Description of an improved Micrometer. \Q7 



in a minute, to do it ; or nearly that. Such a cylinder could 

 be worked bv the descent of ten gallons oF u^ater ten feet m 

 a minute; or, for the whole time, by eighty hogsheads fall- 

 ing the same height. 



But this is a \'ast deal more than could be required, as 

 the fifty people would in eight hours vitiate only three 

 thousand gallons of air, which could be removed by one 

 hundred and fifty strokes of a cylinder, twelve inches dia- 

 meter, with a four-feet stroke, which would not require an 

 expenditure of more than one thousand five hundred gallons 

 of water properly applied, or about twenty-eight hogsheads. 



Holwell, near Tavistock, Feb. 7, 1810. JoHN TaYLOR. 



XXIX. Description of an improved Micrometer. By 

 E. Walker, Esq. 



To Mr. Tilloch. 



Sir, An exact method of taking small angles is a subject 

 of the greatest importance in various branches of practical 

 astronomy and philosophy. For this purpose the micro- 

 meter was invented, which has been constructed of various 

 forms and on different principles ; but even the best of them 

 are very complex in their construction, and consequently 

 too expensive \o be of general utility. 



To obviate these inconveniences Mr. Tiberius Cavallo 

 invented the telescopic mother-of-pearl micrometer," which 

 consists of a small semitransparent slip of mother-of-pearl 

 about the 20th pnrt of an inch broad, and of the thickness 

 of common writing-paper, divided into a number of equal 

 par^s by parallel hues." A full descripiion of this micro- 

 meter and the uses to which it may be applied have been 

 published in the 81st volume of the Philosophical Trans- 

 factions for the year 1791} and in a separate pamphlet piib- 

 4i3hed in 1793.' 



This simple instrument possesses many valuable pro- 

 perties; and when the extremities of ihe object to be mea- 

 jiiir<.'d fall exactly upon two lines ol the scale, it is very ac- 

 curate : but as tbe i'ractional parts of a division on the scale 

 cannot be knf)^vn but by estimation, it falls far short of that 

 T-curacy which is obtained bv other micrometers. 



The micrometer that I have contrived is constructed at a 

 bUi^ll expense, and yet it is very accurate. It consists of a 

 nyiiibtr of parallel hues drawn upon a piece of pbnc glass 

 ** ' with 



