408 Intelligence and Miscellaneous Articles. 



window, while another person below made the experiment as I have 

 indicated. 



I thus observed that the liquid sheet, of a very irregular form and 

 indented on the edges, separates at its edges into numerous solid 

 drops, while the remainder is torn generally into several portions, 

 each of which rapidly closes so as to form a complete hollow bubble. 



My father sees in this phenomenon an argument in support of the 

 vesicular state of the vapour of clouds. In fact one of the principal 

 objections against this hypothesis consists in the impossibility of 

 conceiving how the molecules of gaseous vapour could, when this 

 repasses into the liquid state, agglomerate so as to constitute enve- 

 lopes enclosing air ; but we see now that this immediate agglomera- 

 tion into closed envelopes is not necessary ; it is sufficient that the 

 molecules unite in open plates of any shape and curvature ; each of 

 these plates would then quickly close of itself and give rise to a vesi- 

 cule. Doubtless the production of these plates is not easy to under- 

 stand, but it appeared at least much more admissible than the forma- 

 tion, complete in all respects, of the vesicules. — Bulletin de V Academic 

 Royale de Belgique, ser. 2. vol. xiii. No. 4. 



BY S. M. DRACH, F.R.A.S. 



To the Editors of the Philosophical Magazine and Journal. 



Gentlemen, 

 In Prof. De Morgan's " Notes on Perspective " in the Athenaeum 

 Journal for September 12, p. 335, the author says that the J_(C H) 

 on the base A B= 1 of an equilateral A> as in Euclid I. 1, was pro- 

 posed by A. Durer for a near value of 



360° 

 ch -y-=51° 24' 42"f =0-8677674789, 



which draw (see fig. Euclid I. 1), and call C G. Now C H=%^/3, 

 CG—CH=-001742075 = CHx -00201158. I think this is im- 

 proveable by drawing a straight line C K from C cutting A B in K, 

 so that C K=C G, and K H on line A B 



=VC K 2 — C H 2 = v /-00302039809943 = -05495814859=C H 

 divided by 15*7579072 ; which last number can be gradually approx- 

 imated by 16, ~ t ^ (true to l-45000th). If therefore HK 

 4 oo 



=this aliquot part of C H, and C K be joined, C K=true C G very 

 nearly. If H K, instead of coinciding with H A, be drawn to make 

 angles of 30°, 45°, and 60° with it, C K still remaining =C G, H K 

 will respectively =-00347372, -002461198, -002000109. The 



3 11 



simplest approximation is, however, the above of C K= — .— C H 



and JL to it. The true angle KCH=3° 37' 44"-2, indicative of 

 the closeness of A. Durer's assumption. S. M. Drach. 



London, October 14, 1863. 



Errata. 

 In my Polyhedric Fan paper (April 1863), 



EF=f PD; EPF+3DPE-4CPD=49° 59' 59", &c. 



ftVV. V 



