574 



Mr. J. W. Mallet. 



[June 17, 



other type be substituted for E' it is easy to find its x, for its a. is as 

 area of cups, its f is known, and assuming its x' and computing as 

 before, we get similarly its Ax. He tried five different types and 

 obtained very unexpected results, for lie found that the x varied as some 

 inverse function of the diameter of the cups and of the arms. He 

 gives its values. 



No. 1. Original instrument 12" cups 23*17 arms, 35=1-5880, limit 2-812 



„ 2. Kew 9 „ 24, „ 1*5919, „ 2-831 



„ 3. „ 9 „ 12, „ 1-7463, „ 3-035 



„ 4. „ 9 „ 8, „ 2-1488, „ 4-051 



„ 5. „ 4 „ 26-75, „ 1-8587, „ 3'425 



„ 6. „ 4 „ 10-67, „ 2-5798, „ 4-958 



No. 6 is similar to No. 2, and it might be expected that their con- 

 stants would be equal. The cause of these differences is partly the 

 eddies caused by the cups being more powerful when the arms are 

 short, but still more the presence of high powers of the arm and 

 diameter occurring in the expressions of the mean pressures on the 

 concave and convex surfaces of the hemispheres. In the present state 

 of hydrodynamics we cannot assign these expressions, but we know 

 enough to see that such powers may be present. 



As each type of anemometer has its own constants, the author would 

 suggest to meteorologists the propriety of confining themselves to 

 one or two forms. For fixed instruments he considers the Kew one 

 as good as any, and would wish to see it generally adopted. For 

 portable ones he has no experience except with Casella's 3" cups 6" 

 arms, which he found very convenient ; he has not however deter- 

 mined its constants. Some selection of the sort seems necessary if it 

 is wished to have an uniform system of wind-measures. 



XIII. " Note on the Bearing on the Atomic Weight of Alumi- 

 nium of the Fact that this Metal occludes Hydrogen." 

 By J. W. Mallet, F.R.S. Received June 14, 1880. 



In a recent communication to the Academie des Sciences,* it has 

 been pointed out by Dumas that metallic aluminium and magnesium, 

 as well as silver, may contain sensible quantities of occluded gas, that 

 obtained from aluminium being almost pure hydrogen. 



In a paper on the atomic weight of aluminium read before the Royal 

 Society on the 22nd of April last I gave, among other data, the results 

 of several experiments on the quantity of hydrogen liberated by a 

 known weight of aluminium from a strong solution of sodium hydrate. 



* " Comptes Rendus," 3 Mai, 1880, p. 1027. 



