VI .
" On the Equation of Laplace 's Functions , " &amp;c. By W. F. DONKIN , M.A. , F.R.S. , F.R.A.S. , Savilian Professor of Astronomy , Oxford .
Received December 3 , 1856 .
( Abstract .
) The equation d'u d'u d'u T/ he equation d2 +d+d =0 , when transformed by putting dx2 dky2 ds~2 x= r sin 0 cos 0 , y= r sin sin , z= reos 0 , may be written in the form ( sin )C+d +()sin0)2r(rd + 1 ) } = ; ( 1 ) and if u=uo+ ujlr+ U2r2 + ... + 4 , ^ 4 .
, we find on substituting this value in ( 1 ) , and equating to zero the coefficient of rn , that u , satisfies the equation ( sin 0d )+( )+n(n+ l)(sin0)2 } =0,. .
( 2 ) commonly called the equation of Laplace 's functions .
If we put sill 0 +n cos 0= W , then the equation ( 2 ) may be written dO and t dn2 )+ 0 ; and the operation rw , possesses the following property , namely -nn + --2= , _n1 -(0n-l ) + ( 2n1)2 ; hence it is easily shown , that in general the complete solution of ( 2 ) is Un2= 'nZ~ 1 ** ?
§ 2'Il ?
Uo where uo is the solution of s(in 0w +d ) =o 0 , namely uo/ f(e/ ta-utan - ) +F(e-tan ) ; and the operation wr , Wr-1 ... W2r1 is easily seen to be equivalent to ( sin 0 ) -n sin 0d sin 0 ( This result is compared with that obtained in a different way by Professor Boole ( Camlbridge and Dublin Journal , vol. i. p. 18 ) , to which it bears a general resemblance , but the author has not succeeded at present in reducing the one form to the other .
) In the case in which u , , does not contain i , we have u=C1 + C2 log tan 0 The general expression for a " Laplace 's coefficient " of the nth order , not containing 0 , is therefore ( sin )-n(sin 0 sin 0 .
C ; and if this be called v , when C= 1 , the development of ( 1 -2rcos 0+ r2)-~ is r,2.n , v+vlr+v 22+ + ... +V1 2 +. .
and it is shown that the coefficient of 1-n in the development L. 2 ... n of(1l-2rcos0+r2 ) 2 is ( sin d ( sn 0n 0(sn With respect to the development of ( 1 2r(cos 0 cos 0 ' + sinesin0'cos ?
) + 2 ) -* , it is shown that the coefficient of er cos ip may be put in either of the two forms , .
2 ... ( n i ) 1.2 ... ( n+i ) ( sin 0)-(sin 0')o '( an ) tan 21 2. .
2 ( n ... ( n-i ) .
I. 2 ... ( n ) ( sin 0 ) ) 2~ .
1232 2i : '12 ( sin O)- " ( sin O')nOn-i o ' i(sin 0)2 ( sin o')2i , I. 2 ... ( n-i ) .
I. 2 ... ( n + i ) d where 0 represents the operation sin 0 sin 0 , and the factor 2 is in each case to be omitted when i=O .
( This coefficient is a solution of the equation d( 22 ( sin d+ n(n -t 1 )(sin 0)2_ } U= , of which the complete integral may be expressed in the form ( sin ) sin 0 sin 0 ( sin i)2i(C1 + C2d ( sin 0)-2i-1 ) , at least in the case in which i is an integer not greater than n , for which case this form is here demonstrated .
) If it be assumed that the solution of ( 2 ) , obtained on the supposition that n is an integer , may be extended to the case in which n is a general symbol , it follows that the solution of ( 1 ) will be obtained from it by changing n into r- .
This would give dd u-(sin 0 ) ' ( sin 0 -dsin ) { f(r , e ' tan0 ) +F r , e-/ tan ) } which is easily shown to be equivalent to u=/ f sin 0 Sin e an + sins0 , e0 sin 0 , e-V -'tan ) , where p=r(sin 0)-1 , but p is to be treated as a constant till after all operations .
2 This expression is shown to give known particular integrals , such as ( l--2rcos + r2 ) - , and rn ( sin 0 ) ( sin 0 sin 0 ) ( tan j cos i0 .
It appears probable , therefore , that the generalization of the result obtained for the limited value of n is legitimate ; but the author does not profess to demonstrate this conclusion , believing that the principle of the " permanence of equivalent forms " is not at present established in such a sense as to amount to a demonstration .

