75 .
] On Multiple Contact of Surfaces .
509 jdiately expanding to the same volume as it previously occupied , and a viscosity of the material , which also renders it slow to expand .
Both es causes are , however , rather connected with the effect of the speed the roller on the resistance than with the residual resistance , which , so r as the surfaces are perfectly true and perfectly hard , appears to be ie to the friction which accompanies the deformation , and is hence tiled rolling-friction .
No attempt has yet been made to investigate the laws of rolling-fric-on , although the author hopes to continue the investigation in this direc-on as soon as he has obtained the necessary apparatus .
At the end of the paper attention is called to certain phenomena con-ected with railway-wheels , which it is thought now , for the first time , eceive an explanation .
Thus the surprising superiority of steel rails ver iron in point of durability is explained as being due as much to the act that their hardness prevents the wearing-action , i. e. the slipping , s that it enables them better to withstand the wear .
Also the slipping teneath the wheel explains the wear of the rails in places where brakes \#187 ; re not applied ; and the severe lateral extension beneath the wheel is hought to explain the scaling of wrought-iron rails .
VI .
" On Multiple Contact of Surfaces .
" By William Spottis-woode , M.A. , Treas .
R.S. Received May 24,1875 .
( Abstract .
) In a paper ** On the Contact of Quadrics with other Surfaces , " published in the Proceedings of the London Mathematical Society ( May 14 , 1874 , p. 70 ) , I have shown that it is not in general possible to draw a quadric surface Y so as to touch a given surface U in more than two points , but that a condition must be fulfilled for every additional point .
The equations expressing these conditions , being interpreted in one way , show that two points being taken arbitrarily , the third point of contact , if such there be , must lie on a curve , the equation whereof is there given .
The same formulae , interpreted in another way , serve to determine the conditions which the coefficients of the surface Y must fulfil in order that the contact may be possible for three or more points taken arbitrarily upon it j and , in particular , the degrees of these conditions give the number of surfaces of different kinds which satisfy the problem .
[ In another paper , " Sir les Surfaces Osculatrices " ( Comptes Rendus , 6 Juillet , 1874 , p. 24 ) , the corresponding conditions for the osculation of a quadric with a given surface are discussed .
| In the present paper I have regarded the question in a more general way j and having shown how the formula ) for higher degrees of contact 510 On Non-linear Partial Differential Equations K [ Junel are obtained , I have developed more in detail some special cases i : interest .
For the convenience of the reader , I have in S 1 briefly recapitulat the principal parts of the two papers above quoted .
In S 2 1 have give : at all events , a first sketch of a general theory of multiple contact wi : quadrics ; in S 3 the particular cases of three- , four- , five- , and si pointic contact are discussed ; and in S 4 some conditions for the exis ence of points of four- , five- , six-pointic single ( i. e. not multiple ) conta are established .
Thus far the investigation concerns the contact of quadrics only wit other surfaces .
The concluding part of the paper is concerned with tl corresponding problem for cubics , in which case conditions of possibilil do not arise for simple or two-pointic contact , but are first met with f\lt ; three-pointic contact .
The conditions in question , with some of the consequences , are here given ; and their complexity will perhaps l sufficient justification for not pursuing the subject further in thi direction .
VII .
c\lt ; On the Theory of the Solution of a System of Simultaneou Non-linear Partial Differential Equations of the First Order/ By E. J. Nanson .
Communicated by Prof. Cayley , F.R.S Received June 5 , 1875 . .
( Abstract .
) Given an equation of the form z-f { seva ?
2 , ... .
a\gt ; n+mi av ... anV we obtain by differentiation with respect to each of the n-\-m independen1 variables ocv cc2 , ... .
#n+OT , and elimination of the n arbitrary constant av a2 , ... an , a system of m+1 non-linear partial differential equations o : the first order .
Of this system the given equation may be said to be i " complete primitive .
" Conversely , given a system of non-linear partial differential equations of the first order* it is proposed to determine the conditions which musi be satisfied in order that the system may admit of a complete primitive , and also to examine what kind of solution , if any , exists when the con-ditions above referred to are not satisfied .
The late Professor Boole has given an elegant method of treating a system of linear partial differential equations of the first order ; but the present memoir relates to a more general system , which appears not to have been hitherto considered , viz. to a non-linear system of partial differential equations .
This is here discussed in the two cases\#151 ; first , when the dependent variable z is not explicitly involved in the proposed system ; and , secondly , when z is explicitly involved in the system , the solution of this last case being made to depend upon that of the first-mentioned one .

