44() SUMMARY or ODKRENT RESEARCHES RELATING TO 



Tli( SO can be depended upon. I have never seen one deteriorate by 

 simi^le lapse of time. But these have their disadvantages. They cannot 

 be used with transmitted light, as can scales ruled on glass. Still this 

 difficulty is not insurmountable. I use up to 400 diameters the opaque 

 illuminating objectives made by Bausch and Lomb, which give excellent 

 results. With higher powers, up to 1/18, I have used with satisfaction 

 Prof. Smith's vertical illuminator, with a bull's-eye condenser to concen- 

 trate the light. With a very high power, a 1/18 Zeiss', draw-tube 

 drawn out full length, amj)lifier and high ej e-ijiecin g, I have never yet, 

 on my standard centimeter on speculum metal by Prof. Rogers, been 

 able to see anything but clear sharp edges to the lines, saving now and 

 then a little pit in the metal. Of course, I understand that no practical 

 use can be made of so high a power. I refer to its use simply to show 

 the character of the lines. Any one who has used a glass micrometer 

 with very high powers will agree with me in saying that in this respect 

 they are vastly inferior to those on speculum metal. 



In order not to change the tube-length, when measuring miscel- 

 laneous objects, such as blood-corpuscles, &c., I had Mr. Bulloch make 

 for me an adapter or nose-piece of the same length as my Smith's illumi- 

 nator, also made by him, which I screw on to the front of the tube, in 

 l^lace of the illuminator, when I desire to measure transparent objects. 

 This sort of combination is, in my judgment, the very best that can 

 be used. Metal micrometers have the disadvantage, however, of costing 

 more than scales on glass ; for such a scale should be ruled on a 

 carefully prepared surface, which of course adds to the expense. 



Now as to covering micrometers, in consideration of the disadvan- 

 tages incident to covered scales, I would recommend the use of a scale 

 uncovered. If desired for use with a homogeneous-immersion objective, 

 it can be used with a large temporary cover, which can be held down 

 with a mere dot of mucilage or water, not enough to reach the lines. 

 It should not be rubbed, but may be kept sufficiently clean with a 

 camel's hair pencil. I say sufficiently clean, of malice prepense ; I now 

 think that no one but an amateur with very little experience, will be 

 annoyed by a little dust on a standard when used with a dry objective. 

 If it becomes too thick, it can be removed with a camel's hair pencil. 

 If used with an immersion objective, of course the top of the temporary 

 cover should be clean. I find a little dust a real convenience, as facili- 

 tating the finding and focusing of the lines. A really fastidious person 

 f-hould use " Centimeter A " for a time. Its surface, the last time I saw 

 it, was in places seamed and furrowed, like the track of a glacier. But 

 enough of it is perfect for any sort of use, and its lines cannot well be 

 excelled. Its correction for total length is very small, and its second 

 mm. has practically no error. 



Of course a micrometer in its ultimate subdivisions, such as are 

 usually used in determining the value to be assigned to one division of 

 the eye- piece micrometer, should have an error so small as to be practi- 

 cally insensible, or its error should be well determined, I have never 

 yet seen, nor do I ever expect to see, a scale in every part absolutely free 

 from error. I undertake to say that such a scale cannot be made by any 

 living man, but the absolute and relative errors of a scale can be deter- 

 mined within very narrow limits, and a scale can be made, the errors of 

 whose ultimate subdivisions are practically insensible. Such a micro- 

 meter is practically perfect. In a future communication, should the 



