XI .
Of the Fluents of Multinomials , and Series affected by radical Signs , which do not begin to converge till after the second Term ; in a Letter from T. Simpson F. R. S. to W. Jones Esq ; V. P. R. S. ALTH O ' the Application of infinite Series , and the Quadrature of the conic Sections , to the inverse Method of Fluxions has exercised the Pens of the most able Mathematicians , and produced many curious and useful Discoveries , yet nothing has been hitherto given , that I know of , whereby the Fluents of radical Multinomials and Series , which do not begin to converge till after the second Term , can be determined , so as to be of Use in the Solution of Problems : The common Method , by expanding the given Expression , being , you know , altogether impracticable in this Case .
Presented May 26 .
1748 .
The Consideration of which induced me to draw up the following Paper ; which I humbly beg Leave to lay before you , who are so good a Judge of the various Improvements which this Subject has from time to time received .
What most encourages me to hope this little Essay will meet with your Approbation , is , that it is not merely an abstracted useless Speculation , but may be applied to good purpose in many difficult and important Enquiries into Nature ; whereof I have put down one or two Instances , and shall further take the Liberty to observe here , that most of the lunat Equations , given by Sir Isaac Newton , are only such Approximations as may be exhibited by the first Term of a Series derived by the Method here delivered .
Proposition .
The Fluent of a+c zn mxz pn-1 z being given ( either in algebraic Terms , or from the Quadrature of the Conic Sections , &amp;c. ) it is proposed , by means thereof , to approximate the Fluent of a+ cxn + dx2n + ex3n + fx4n &amp;c. mxxpn-1 x ; supposing the Series not to converge till after the second Term .
Make czn = cxn + dx2n + ex3n~c .
and let Q be the given Fluent of a+ czn m xz pn-1 z , answering to any proposed Value of x : Moreover let y= xpn , or y1/ p = xn , and let this Value of xn be substituted in the first Equation , and it will become czn = cy1/ p + dy2/ p + ey3/ p ~c .
whereof the Root y being extracted , we shall ( by making R= pd/ c , S=pp+ 3/ 2 x d2/ c2 pe/ c , T= -p p+4 p+5/ 6 x d3/ c3 + p. p+4/ 1 x de/ c2 pf/ c ~c .
) have y ( xpn ) = zpn + Rz pn+n + Szpn+2n ~c .
whence we also obtain x pn-1 x=z pn-1 z+ p+1/ p x Rz pn+n-1 z + p+2/ p x Szpn+2n-1 z ~c .
Let this Value , with that of cxn + dx2n + ex3n ~c .
( above given ) be now substituted in the proposed posed Fluxion , and it will become a+c zn m xz pn-1 z+p+1/ pxRzpn+n-1 z+p+2/ pxSzpn+2n-1 z &amp;c. Moreover , let v denote the Place , or Distance , of any Term , of this Expression , from the first ( exclusive ) then the Term itself ( drawn into the common Multiplicator ) will be denoted by a+ czn m xp+v/ p x Az pn+vn-1 z ; and the Fluent thereof will be truly expressed by p+1/ p+m+1 x p+2/ p+m+2 x p+3/ p+m+3 x ... p+v/ p+m+v x a/ -c vxAQ+ p+v .
A/ p x a+czn m+1 xzpn-n/ p+m+vxnc into z un p+v-1/ p+m+v-1 x az vn-n/ c +p+v-1 p+v-2/ p+v+m-1 .
p+v+m-2 x a2 zvn-2n/ c2 ~c .
continued to as many Terms as there are Units in v. Wherein let v be expounded by 1 , 2 , 3 ~c .
successively , and R , S , T , ~c .
by A respectively : By which means the Fluent of the whole Expression will be obtained .
Corol .
1 .
Because the Fluent of the general Term , when the Multiplicator a+cz m+1 becomes = o , is barely = p+1/ p+m+1 x p+2/ p+m+2 x p+3/ p+m+3 x ... .
p+v/ p+m+v x av/ x AQ the Fluent of the whole Expression will , therefore , in this Case be truly defined by Q~1p+ 1 .
R a/ p +m+ 1 .
C+ p+1 .
p+2 .
S2 a/ p +m+ 1 .
p+m+ 2 .
C2 p+1 .
p+2 .
p+3 .
Ta3/ +m+1 .
p+m+2 .
p+m+3.c3 ~c .
Where Q denotes the Fluent of a+czn m ~z pn-1 z , when zn=a/ -c .
Corol .
2 .
But , if m+1 and p be , each of them , the Half of an odd affirmative Number , and P be taken to denote the Periphery of a Circle whose Diameter is Unity , and -c be put =b , then the Valne of Q ( or the Fluent of a-bzn m~ zpn-1 z , when zn=a/ b ) will be = ap+mp/ nbp ~ 1 .
3 .
5 .
7 .
~c .
( to p-1/ 2 Factors ) ~ 1 .
3 .
5 .
7 ~c .
to ( m+1/ 2 Factors)/ 2 .
4 .
6 .
8 .
10 .
12 ~c .
( to p+m Factors ) Therefore the Whole , required , Fluent , of a-bxn + dx2n + ex3n ~c .
m ~xpn-1 x is , in this Case , equal to the Product of that Expression into the following Series , 1+p+1 .
Ra/ p +m+ 1.b + p+1 .
p+2 .
Sa2/ p+m+1 .
p+m+2 .
b2 ~c .
Wherein R is to be taken = pd/ b , S=p .
p +3/ 2 ~ d2/ b2 +p e/ b , T= p. p+4 .
p+5/ 6 ~ d3/ b3 + p. p+1/ 1 ~d e/ b2 +p f/ b , ~c .
according to what is above specified .
The Use of what has been delivered above will , in some measure , appear from the Solution of the two following Problems , which I shall subjoin as Examples thereof .
The first is ; To find the Time of Oscillation in the Arch of a Cycloid , in a Medium resisting according to the duplicate Ratio of the Velocity .
Let A denote the whole Arch of the SemiCycloid , or the Length of the Pendulun , a the the Arch described in the whole Descent , and x any variable Part thereof described from the Beginning of the Descent ; and let the Density of the Medium be , every-where , as 1/ b : Then the Fluxion of the Time will be found = a-1 + 2a/ h ~ x/ 2-2 x2/ 2.3h + 4x3/ 2 .
3 .
4h2-8x4/ 2 .
3 .
4 .
5h3~c .
-1/ 2~ 2A 1/ 2 ~x1/ 2 x * : which being compared with a-bxn+dx2n+ex3n ~c .
m~ xpn-1 x ( vide Corol .
2 .
) we shall , in this Case , have n=1 , m=-1/ 2 , p=1/ 2 , a=a , b=1 ~ 2a/ h~1/ 2 , d/ b=2/ 3h , e/ b=-1/ 3h2 , f/ b=2/ 15h3 ~c .
Whence R= 1/ 3h , S= 2/ 9h2 , T= 8/ 45h3 ~c .
Also a p+m P/ nbp ~ 1 .
3 .
5 .
7 ( p-1/ 2 ) ~ 1 .
3 .
5 .
7 ( m+1/ 2)/ 2 .
4 .
6 .
8 .
10 .
( p+m ) = P/ b1/ 2 , and 1+ p+1 .
Ra/ p+m+1 .
b+p+1 .
p+2 .
Sa2/ p+m+1 .
p+m+2 .
b2 ~c .
= 1+a/ 2 b h+ 5a2/ 12b2 h2 * The Investigation of this , and the Fluxion in the following Example , are both given in my Essays .
+5a2/ 12b2 h2+7a3/ 18b3 h3 ~c .
Whence we have 2A 1/ 2 ~ P/ b1/ 2 ~1+ a/ 2bh + 5a2/ nb2 h2 , ~c .
for the Time of one Vibration of the Pendulum ; which , by substituting 1+ 2a/ h ~ 1/ 2 for its Equal b , ~c .
becomes P A1/ 2 ~1*+ a2/ 6h2 2 a3/ 9h3 ~c .
From which it appears , that the Effect of the Resistance on the Time of Vibration , in small Arches , is nearly in the duplicate Ratio of those Arches .
Sir Isaac Newton ( from whom it is impossible to disagree without being under some Apprehensions of a Mistake ) has , indeed , given a very different Solution to this Problem ( in Princip .
Prop. 27 .
B. 2 .
) .
But as the Conclusion here brought out exactly agrees with what I have elsewhere given , by a different Method , I have great Reason to believe I have no where fallen into an Error .
The second Example I shall give as an Illustration of the foregoing Method is , To determine the Apside Angle ( or the Angle of the two Apses at the Center ) in an Orbit described by means of a centripetal Force , which varies according to any Power of the Distance .
In order to which , let the Velocity of the Body at the higher Apse be to that whereby it might describe a Circle at the same Distance from the Center , in the given Ratio of p to Unity ; also let that Distance be denoted by Unity ; and , supposing z to denote any other Distance , let the centripetal Force be universally expressed by zn .
Then the Fluxion of the Angle at the Center will be expressed by p z/ z~ p2 + 2/ n+1 x z2 p2 2z/ n+1n+3 Put a= 1p2 , v= n+3/ 2 and x=1 z~ , and it will become 1/ 2~ 1ax ~x/ 1 xx~ax+1vx1 x~~/ 1 v= 1/ 2 1a 1/ 2 into av x/ 2 + v. v-2/ 2 .
3x x2 v. v-2 .
v-3/ 2 .
3 .
4x x3 &amp;c. ~ 1/ 2 x x -1/ 2 x+x 1/ 2 ~x +x 3/ 2 ~x &amp;c. Now , to find the Fluent of the first Term hereof ( drawn into the general Multiplicator ) or av x/ 2 + v. v-2/ 2.3 .
x x2 &amp;c. ~ -1/ 2xx -1/ 2 x , we have ( as before ) n= 1 , m= 1/ 2 , p= 1/ 2 , b= v/ 2 , d/ b = v-2/ 3 , e/ b =-v-2 .
v-3/ 3.4 , &amp;c. Also R=v-2/ 6 , S=v-2 .
4v-5/ 72 ; and consequently the Fluent itself ( when the Body arrives at the lower Apse ) =P/ ~ 1/ 2 vx 1+ v-2/ 2v xa+ 5 .
v-2 .
4v-5/ 48v2 x a2 + 7 .
v-v .
16v2 37v + 22/ 6 .
48v3 &amp;c. After the same manner the Fluent of the the second Term will come out = P/ ~ 1/ 2 vx a/ v + 5 .
v-2/ 4v2 x a2 + 3 .
v- .
~-3/ 48v3 x a3 &amp;c. that of the third = P/ ~ 1/ 2 vx 3a2/ 2v2 + 35 .
v-2/ 12v3 x a3 &amp;c. &amp;c. &amp;c. Whence , by collecting these several Fluents together , we have P/ ~ 1/ 2 vx 1+1/ 2a + 20v2-5v+2/ 48v2 x a2 + 112v3-63v2+42v-8/ 6 .
48v3 xa3 &amp;c. for the Fluent of the whole Expression : And this , drawn into 1/ 2 x1 a/ 2 a2/ 8 &amp;c. ( =1/ 2x~1-a~1/ 2 ) will be =P/ ~ 2vx1* v-2 .
2v-1/ 48 x a2/ v2 + v-2 .
2v-1~2/ 72 x a3/ v3 &amp;c. =P/ ~ n+3x1 *n-1 .
n+2/ 24 x a2/ n+3~2 + n-1 .
n+2~2/ 18 x a3/ n+3~3 &amp;c. which , in Degrees , gives 180/ ~ n+3x 1~ +n-1 .
n+2/ 24 x a2/ n+3~2 +n-1 .
n+2~2/ 18 x a3/ n+3~3 &amp;c. for the true Measure of the Angle required .

