Fourth Series. — Metallic Gratings. 113 



voire is fine compared to the interstices, the interstices are 

 pretty nearly rectangular and equal-sided. But this is not the 

 case in most manufactured wire-gauze. One set of wires is 

 nearly parallel and straight, but not so the set interlaced with 

 the former, which do not generally make their intersections at 

 right angles, and hence, universally, the interstices are some- 

 what smaller than a calculation proceeding upon the number 

 of wires per inch, and their diameter would give. Distrust- 

 ing my own observations, I put three specimens of wire-gauze 

 into Mr. John Adie's hands, requesting him to determine the 

 mean diameters and intervals of the wires. With a very ac- 

 curate micrometer he determined 14« values for each of these 

 quantities in both directions. From these data the proportion 

 of the interstices to the whole surface of each grating is easily 

 calculated, and the results are given below for three sorts of 

 gauze of which I had previously determined the permeability 

 for heat. 



Micrometric Measurement of Wire-Gauze. Unit of Mea- 



sure = 



i_ 



T2T0 



inch. 



33. The numbers in the last column (computed on the 

 supposition of the interstices being geometrical rectangles) are 

 to be compared with the following experimental determina- 

 tions of the proportion of incident heat transmitted by these 

 gratings. 



Proportions per 100 of Incident Rays of Heat transmitted by 

 Wire-Gauze. 



Wire-Gauze. 



No. 1. (57) . 

 No. 2. (92) . 

 No. 3. (129) 



Locatelli , with 

 Glass. 



32-5 



460* 



30-5 



Locatelli. 



32 



Dark Hot 

 Brass. 



33-5 



44-7 f 

 30 



Hot Water. 



29-7 



* Two such gratings superimposed, so that the wires formed angles 

 of 45° with one another, gave for the per-centage of transmission 20-7. 

 The square root of this, or effect due to each grating, is 45-5, or almost 

 the same as the number in the text. 



t Two superimposed gratings gave 21-2 per cent., or 46 for each 

 system separately. 



Phil. Mag. S. 3. Vol. 19. No. 122. Aug. 184-1. I 



