646 REPORT— 1892, 



taining which part of the flame took part in the production of the spectra ohserved 

 when metallic salts are introduced into non-luminous flames. The main conclusion 

 derived was, that spectral eflects are dependent, in the cases investigated, on 

 radiations produced during chemical change, and not to molecular vibrations 

 induced by external heating. An experiment was shown in which a spray of a 

 solution of a copper salt was introduced into the two separated cones of a non- 

 luminous flame. Decomposition of the salt took place in the inner and hotter 

 cone without the production of a spectrum, whilst the products of decomposition 

 on reaching the outer flame tinged it a brilliant green. This spectrum was thus 

 shown to depend on chemical action, and not merely on external heating, the 

 action in this case being evidently the formation of copper oxide. The formation 

 of copper chloride, which could be efi^ected by introducing a thread of asbestos, 

 moistened with hydrochloric acid, into the outer cone, gave a beautiful blue 

 colour. Gold chloride similarly employed was decomposed also in the lower cone 

 without producing a spectrum. The separated gold passed into the upper flame, 

 which was faintly tinged ; but if a little hydrochloric acid were introduced, a 

 beautiful green colour (due to gold chloride) was observed. The action between 

 gold and hydrochloric acid at a temperature above that at which gold chloride 

 dissociates, and below that at which hydrochloric acid dissociates, seemed highly 

 anomalous. The author did not wish at present to draw any final conclusions, but 

 considered that his experiments disclosed the necessity for a thorough study of the 

 chemical reactions taking place when metallic salts are producing spectra in non- 

 luminous flames. 



Department II. — Mathematics. 



1. Models of Mathematical Surfaces. By Professor A, Crum Beown, 



F.B.S., F.B.bl.E. 



The models exhibited were : — (1) Two in plaster of Paris of surfaces with a 

 cubic and biquadratic equation respectively. (2) Of a surface described by a 

 generating line always at right angles to a tangent at a point on a fixed circle. 

 The point moves round the circle at an angular rate twice that at which the 

 generating line turns about the tangent, so that the two sides of the surface are 

 continuous. The surface contains, besides the generating lines, two straight lines, 

 viz., the axis, i.e., the line through the centre of the circle at right angles to its 

 plane, and the intersection, i.e., the straight line in which the surface cuts itself. 

 (3) A number of models showing how three interlocked surfaces may be applied to 

 (a) a cylinder and (6) an anchor ring. 



2. Interim Report of the Committee on Tables of Mathematical Functions. 



The Committee intend to present a report at the next meeting of the 

 Association. 



3. Interim Report of the Committee on Tables connected with the Pellian 



Equation. 



The Committee intend to present a report at the next meeting of the 

 Association. 



4. Note on the GUssettes of an Ellipse and of a Hyperbola, 

 By Professor P. G. Tait, Sec.B.S.E. 



The author exhibited metallic discs constructed for the purpose of illustrating 

 the singular theorem that the curve, described by any point in the plane of an 

 ellipse, which slides between rectangular guides, can be described by a definite 



