of the Red Admiral Butterfly, and the Lejnsma Saccharina. 61 



A remarkable deterioration in definition occurs with scales 

 attached to the slide and laying below those adherent to the cover. 



In order to measure the diameter of these cross trios of spherules, 

 an eye-piece micrometer was now used to examine the thousandths 

 of an inch on a stage micrometer. The divisions in the eye-piece* 

 were 200 to the inch. One stage thousandth exactly measured 

 13*5 divisions; the power employed was 675 = 1350 -f- 2. The 

 spherules appeared to be just four to the width of one division of 

 eye-piece. The diameter of one spherule therefore = £th of these 

 divisions. 



The spherules are therefore about l-44000th of an inch, 

 set close together in contact. Now it is singular that although 

 Nobert's Band VII. on the new plate contains about 



45,000 divisions to the inch, 



they are very much easier to be seen than these spherules separated 

 from centre to centre by almost the very same interval. Indeed, 

 in the case of the spherules in contact there are only minute 

 notches OCOO, whilst in Nobert's bands there is a clear space between 

 the lines ; and this notch is much less than the spaces of Nobert's 

 lines of Band Vll.f 



I think that those of our Fellows who will take the trouble to 

 examine the spherules of the Bed Admiral scales, will quickly 

 discover that (with an inch and a half object-glass condenser) on 

 a cloudy day it is only by their very best glasses they can 

 detect the cross rows of beads in the intercostal spaces. Perhaps 

 the rationale of the easier visibility of Nobert's lines 50,000 to the 

 inch, as compared with equivalent spherules in close contact, will 

 be more apparent by means of a diagram.^ 



Let A B, A B be a pair of Nobert's lines of Band VIII., about 

 50,000 per inch ; C G, the centres of two spherules, whose diameter 

 is the same, viz. l-50000th. Now the visibility of these spherules 



* This eye-piece is described at page 5 of the Inst number of Journal. When 

 the " lens-micrometer" is divided 100 per inch, the power 675 is indicated at sight 

 bv the stage micrometer, but at 200 per inch, it would read, of course, 1375 instead 

 of 675. 



t Tolles' immersion Jgth with a power of 800 showed the 8th band with 

 central light, which rates at about 50,000 per inch. Wales' -i-th with a power of 

 475 showed them with oblique sunlight : a fine performance. 



X The letters A A, B B in the diagram should have been engraved at points 

 close to the lines touching the circles ; moreover, the fine lines drawn by the en- 

 graver are merely intended as a shading, A C B in each rate representing tiro cm- 

 secutive single lines of a band. 



F 2 



