]\gt ; 14 Mr. W. H. L. Russell on Linear Differential Equations .
[ Nov. 21 , tion was not examined by means of the stethometer ; but from Mr. Le Gros Clark 's experiments it would appear that they have but little influence in forced inspiration .
With the intrinsic tboracic muscles the case is different .
It is probable that , with the exception of the levatores costarum , each of these muscles has some further power beyond that of simply raising or depressing the ribs .
The triangularis sterni is more a constrictor of the anterior part of the thorax .
The xtercostals have very various actiollS assigned to them ; but when all the evidence of Traube , Sibson , and others is considered , it seems most probable that the six upper ribs are raised by the external intercostals , and depressed by the internal muscles ; but the ribs are not rigid bars , and hence these muscles have a further action upon them .
From careful experiments upon a model with elastic vulcanite ribs , and elastic bands stretched between them in the direction of ( a ) the external intercostals , ( b ) the internal intercostals , and with now ths upper ribs fixed , now the lower , it was comcluded that whilst modifications were introduced into their action by the last-named condition , yet that without doubt the tendency of the external intercostals was , 1st , to draw the ribs upwards ; 2nd , to separate their anterior ends ; 3rd , to straighten them .
On the other hand , the ction of the internal intercostals was , 1st , to draw the ribs vnwards ; 2nd , to bring their anterior extremities nearer together ; 3rd , to bend them inwards .
These results were explained by resolving the forces of these muscles in the directions ( a ) along the ribs , ( b ) at right angles to them .
The diaphragm may also be considered to have an important action in bending the lower ribs .
Evidence can be adduced showing the value of the stethometer used in these inques , both in physiological studies and in medical practice .
III .
" " On Linear Differential Equations.\ldquo ; \mdash ; No. By W. H. L. F.R.S. Received July 1872 .
We now consider ]inear differential equations which are satisfied by the roots of an algebraical equation admitting of explicit solution .
To determine in what cases the ]inear diiferential equation is satisfied by assuming when X , are rational functions of .
We shall commence by 872 .
] Mr. W. H. L. Russell on Linear Differential .
15 supposing them also entire functions of .
At present we shall confine ourselves to the linear equation of the second order , The most general form of solution which this equation admit will be .
For if put , we should have three rticular integrals correspondin to the three cube roots of being stlpposed fractional .
Substituting in the proposed tion , we shall find , after making some reductious , ; . . . .
( 1 ) from whence we have , . . . . .
( 2 ) . . . . . . . .
( 3 ) Let This will give us For if we assumed either X or to be of higher dimensions , we should obtain , on equating the coefficients of the powers of ( x ) in 2 and 3 to zero , more equations than the total number of constants in the differential equation and assumed solution .
The number of disposable constants is eleven .
If we assumed X of two dimensions or of dimensions , we should have more than eleven equations .
Now equate the coefficient of the highest power of ( x ) in ( 2 ) to zero , and we have which determines independently of the constants in X and Y. Similarly , from ( 3 ) , This of course becomes an equation of condition betweeJl the coefficients of the given diflerential equation , when we substitute the value of ( n ) we have found from the last equation .
To determine , we proceed as foUows:\mdash ; In equation ( 3 ) put successively equal to ths three roots of the equation Then we have three equations , linear in , from which these qu 16 .
W. H. L. Russell on Linear Differential Bquations .
[ Nov. tities may be determined .
But putting in equation ( 1 ) , the equation becomes Hence , substituting for and the lues we have just obtained , we have a linear equation to determine ; hence determined .
Now , suppose to be such that we may assume ; then the terms of equation ( 2 ) , which contain the two highest powers of , are linear in , or we may divide it out in equating their coefficients to zero .
Hence we may determine at once .
Put for successively in ( 2 ) four roots of the equation ; then we have four equations ]inear in , whence we have a quadratic equation to determine , and then are known .
We have beside a number of equations of condition .
Now suppose to be such that we may have ; then we shall find that the terms which contained the three highest powers of ( x ) are linear in ; consequently , and therefore X , may be determined at once .
X being known , we may obtain a series of equations easily by which the constants in may be determine .
But we may adopt a somewhat different method , which be useful in many of this nature .
Resuming the equation equation ( 1 ) shows that must vanish for every quantity which causes to vanish , and equation ( 3 ) that the quantity must also vanish for every quantity which causes to vanish .
Let .
* If he terms which contain the highest powers are linear in ; and therefore can be determined at once , commenoing with If the terms which contain highest powers are linear in ; and foro may be determined at once.\mdash ; W. .
L. R. , Nov. 21 , 1872 .
187 .
W. H. L. Russell on Linear Differentiat Equations .
17 Then must be one of the quantities suppose must have ; hencc , if be lineal .
equation , admitting a solution of the form .
to ; every qtityhich m nish ; every quantity which . .
to yauish make v. This pl.inciple is of of the greatest use in determiming the form of this function .
The most general form of irrationitl solution for it linear equation of the third order be ; for one of the fourth order we shall have where is supposed to be all these the tion of be independent of constants in X , , Z. For if we expand the proposed forms of in descending -ers of , substitute in the lalion , eqnate the coefficient of the of ( x ) to zero , } obtain an equation to not ] lvolving these constants .
If X , , . . .
are rational fractioJls , this method not apply ; but since the factors of the denominators of these frnctions , not rding .
powers , must also be factol.s of proceed Let be oue of the fiactors of , substitute.x in the diflerential equation , and let be the test n of obtained by substitutiug in the differential equation , and equating the vest term of the result to zero .
, let { let be the test negative of , obtained by substiluting in like manner .
Then , if lve put obtain au equation whose solution will be free from negntive factors .
VOL. XXI . .
W. H. L. Russell on Linear Differential Equations .
[ Nov. We have seen that if , when substituted for , satisfies a linear tion , is is , it remains the same function of , and the constants involved in the equation whatever numerical we to assign to these coIlstants , on the supposition that and rational fUmctions of .
shall call this the " " principle of d proceed to show that it is true when is irrational .
Let be equation ; to find the condition that it may be satisfied by is .
The tional qantit in must be of the fol.m vise the equation would admit of more than two particular integrals .
Ioreover the ffictors of X must be factors of We will suppose .
Hence we may assume ; for if we expalld in descending powels of , and substitute the resulting of the differential shall easily see that it can have no power higher ] the first .
, and are of course constants to be determined , and the double sign of the radical will give rise to four equations of two , which imply equation of coIldition .
We now substitute , where is ffigebraical function of , in the linear differential equation , and thus obtain The of the two highest powers of ( x ) must be equated to zero , that is to say , the coefficisnts of the two highest powers of the multiphel ' of must be equated to zero .
The followin of that multiplier will be suffioient for our purpose : or 1872 .
] Mr. W. H.L. Russell on Linear Differential Equaiions .
19 Hence , the double of , . . . . .
( 1 ) , . . . .
, ( 3 ) From the first two ions , ; and , bstituting in the fourth , Consequently we shall have the in the diffel.ential tion must tisfy the condition from the third equation , It is easy to see that similar proof will apply in other cases .
We n-ill lecapitulate our results .
The solutions of differential equations , when they can be ined in fmite terms , must be cither , exponential , or arithmic .
If , they must be either or irrntional .
The method of finding the rational algebraical solutions has been oul.ed the present to show how the solufious be for tions of the second order .
proved that diff .Jlti 1 not dmit of olutiol of the ; still less will it admit of solution of the form when is of .
The exponential solution must be of the where is an riant , that is , function which remains umchanged whatever algebraical value we may be able to assign to P. This invariallt enables us to reduce linear differential equations containing functions to other ions i ] fullctions only .
We olutionS of linenr differential equations must in general be of the form ] is lgebraical function satisfying the proposed linear cquntioJl , and be deduced from another linear eqnation , which have } } how fo construct .
A great remains to be done ; ertheless I think we not obsourely the true fouudations on which the solution of differential equations in finite terms must of necessity rest .

