on l\Icrino Sheep and IFool. 2 29 



Neiiher is its inaccuracv less than its difficulty; for if, after we have succeeded 

 in making one edge of a filament exactly correspond with one side of a square in 

 the micrometer, the diameter of the filament were in the smallest degree less or 

 greater than the -—-^ of an inch French, the other edge of the filament would not 

 reach, or would reach beyond, the next parallel line of the micrometer, and we 

 should be left merely to gucfs at the diameter. 



If it were attempted to remedy llii< defect bv a micrometer more minutely sub- 

 divided, those sub-divisions would be proportionably inaccurate; and the difficulty 

 of adjustment before stated would remain in full force. 



These methods, therefore, promise little success ; and the inaccuracy of the results, 

 which I shall hereafier point out, strongly confirms the opinion that they are 

 inadequate. 



What mode Mr. Luccock has employed in order to apply a scale to a magnifier^ 

 he has not informed us ; but from the errors of mensuration into which he also ap- 

 pears to me to have fallen, 1 think it probable that his method is still far from being 

 correct. 



Under these difficulties I applied for assistance to a man, whose genius and science 

 are only equalled by the warmth of his friendships, and the readiness with which 

 he communicates bis knowledge to those who are desirous of sharing it. Dr. 

 Herschel advised me to make use of the method which lie has described in 

 the account of his lamp-micrometer, published in the Philosophical Transactions 

 for 1782 ;* and also in a letter to Sir Joseph Banks, contained in the same volume. t 

 Availing myself of these suggestions, I found, that, after some experience, the re- 

 lative diameters of any minute objects could easily and with great precision be 

 ascertained by the following method. An object of a known diameter being placed 

 on the stage of the microscope in a strong light, and a piece of white paper spread 

 horizontally beneath it, look through the tube at the object with both eyes open, 

 and you will see its image projected on the paper below, which you may then mea- 

 sure with a pair of compasses, and take off the aperture on a diagonal scale minutely 

 subdivided. Divide this magnified image by the known diameter of the object, and 

 you will find the magnifying power. This being found, place on the stage the 

 object, the diameter of which is sought, and having measured with the compasses, 

 as before, the diameter of its image projected on the paper below, divide that dia- 

 meter by the magniiying power, and the quotient will be the real magnitude required: 



* Tagf 165. t Page 174. 



3 V 



