nS- IMAGE EVALUATION TEST TARGET (MT-3) w ^ A ^ ^ '^ ^0 ^ 1.0 1.25 ly ■^" ■■■ la y^ |22 m ii& 12.0 |U 11.6 oS. '^ y^ 7 ffiotographic Sciences Corporation 33 WiST MAIN STRUT WIBSTIR,N.Y. 14SS0 (716)«72-4S03 # ^ ;V 4 '^'^^ 4^ CIHM/iCMH Microfiche Series. CIHIVI/iCMH Collection de microfiches. Canadian Institute for Historical Microreproductions / Institut Canadian de microreproductions historiques Ttchnical and Bibliographic Notaa^'Notas tachniquaa at bibliographiquaa Tha Instituta haa attamptad to obtain tha baat original copy availabia for filming. Faaturaa of thia copy which may ba bibliographicaily uniqua, which may altar any of tha imagaa in tha raproduction. or which may aignificantiy changa tha uaual mathod of filming, ara chaekad balow. □ Colourad covara/ Couwrturo da coulaur r~| Covara damagad/ D D Couvartura andommagte Covara raatorad and/or laminatad/ Couvartura raataur^ at/ou pailiculAa □ Cover titia missing/ La titra da couvartura manqua Colourad mapa/ Cartaa gtegraphiquaa tt coulaur □ Colourad ink li.a. othar than blua or black)/ Encra da coulaur (i.a. autra qua blaua ou noira) rn Colourad plataa and/or iliuatrationa/ D D D D Planchaa at/ou iliuatrationa 9n coulaur Bound with othar matarial/ Rali4 avac d'autraa doeumanta Tight binding may cauaa shadowa or distortion along intarior margin/ La f liura sarria paut causar da I'ombra ou da la diatorsion la tong da la marga intiriaura Blank laavas addad during rastoration may appaar within tha taxt. Whanavar possibia, thasa hava baan omittad from filming/ II sa paut qua cartainas pagas blanchas ajoutias lors d'una rastauration apparaissant dana la taxta, maia. lorsqua cala itait poasibla. caa pagaa n'ont paa *ti filmiaa. Additional commanta:/ Commantairas suppiimantairas; Th to L'Inatitut a microfilm* ia maillaur axamplaira qu'il lui a iti possibia da sa proeurar. Las details da cat axamplaira qui sont paut-4tra uniquas du point da vua bibliographiqua. qui pauvant modifior una imaga raproduita. ou qui pauvant axigar una modification dans la mithoda normala da filmaga sont indiqu^a ci-daaaoua. □ Colourad pagaa/ Pagaa D Pagaa da coulaur Pagaa damagad/ Pagaa andommagiaa Pagaa raatorad and/oi Pagaa raatauriaa at/ou pailiculAas Pagaa discolourad, stainad or foxai Pagaa dicoiorias. tachatAas ou piqutea Pagaa datachad/ Pagaa ditachias Showthrough/ Tranaparanca Quality of pr\r\ Quaiiti inigala da I'imprassion Includaa supplamantary matarii Comprand du material suppl4mantairn Only adition availabia/ Saula Mition disponibia r~l Pagaa damagad/ pn Pagaa raatorad and/or laminatad/ p~1 Pagaa discolourad, stainad or foxad/ r~| Pagaa datachad/ rri Showthrough/ rn Quality of print varias/ r~n Includaa supplamantary matarial/ pn Only adition availabia/ Pagaa wholly or partially obscured by errata slips, tissues, etc.. hava been refilmed to ensure the best possible image/ Lea pages totaiement ou partieilement obscurcies par un fauillet d'errata. una peiure, etc.. ont iti filmies i nouveau da fapon A obtanir la mailleure imaga possible. Th pa of fill Oi hi th ah ot fir al( or Tt ah Tl wl M di an bi ri| rm mi This item is filmed at tha reduction ratio checked below/ Ce document est filmA au taux da reduction indiqui ci*daaaoua. 10X 14X 18X 22X 12X 16X aox 2SX aox 24X 28X n 32X TIm copy fllm«d her* haa Immi raprodiie«d thankt to tho gonorotlty of: DouglM Library Quaan'a Univarsity L'axamplaira fNni4 fut raproduK grAoa i la g4n«roalt« da: Douglaa Library Quaan'a Univaraity Tha imagaa appaaring tiara ara tha baat quality poaaibia eonaidariiig tha condition and iagiblHty of tha original copy and In kaaping with tha filming contract apaciflcatlona. Original copiaa in printad papar covara ara flimad baginning with tha front covar and anding on tha laat paga with a printad or iliuatratad impraa- aion, or tha back covar whan appropriata. All othar original copiaa ara flimad baginning on tha f irat paga with a printad or illuatt'atad impraa* aion, and anding on tha laat paga with a printad or iliuatratad impraaaion. Laa imagaa aulvantaa ont 4t4 raproduitaa avac la plua grand coin, compta tanu da la condition at da la nattat* da i'axamplaira film4, at an conformit* avac laa conditions du contrat da fllmaga. Laa axamplairaa originaux dont la couvartura an paplar aat Imprimte aont fllmte an commandant par la pramlar plat at an tarmlnant aoit par la darnlAra paga qui comporta una ampralnta dimpraaalon ou dlHuatration, aolt par la aacond plat, aalon la caa. Toua laa autraa axamplairaa originaux aont fllmAa an comman9ant par la pramMra paga qui comporta una ampralnta dimpraaalon ou d'liiuatration at an tarmlnant par la damlAra paga qui comporta una talla ampralnta. Tha laat racordad frama on aach microficha ahall contain tha aymboi -^ (moaning "CON- TINUED"), or tha aymboi V (moaning "END"), whichavar appliaa. Un daa aymbdaa auh^anta apparaltra aur la darnlAra Imaga da chaqua microficha, aalon la caa: la aymbola — ► algnlfia "A 8UIVRE", la aymbola V algnlfia "FIN". IMapa, plataa, charta, ate, may ba flimad at diffarant raduction ratioa. Thoaa too larga to ba antiraly included in ona axpoaura ara flimad baginning in tha uppar laft hand corner, loft to right and top to bottom, aa many framaa aa raquirad. Tha following diagrama liluatrata tha mathod: Laa cartaa, planchaa, tabiaaux, ate, pauvant Atra filmte A daa taux da rMuction diff Aranta. Loraque la document oat trap grand pour Atra raproduit an un aaui clichA, II aat fllmi i partir da i'angia aupArlaur gaucha, da gaucha A droita, at da haut an baa, an pranant la nombra d'imagaa nteaaaaira. Laa dlagrammaa auivanta liluatrant la mithoda. 1 2 3 1 2 3 4 5 6 J QUEEN'S COLLEGE MATHEMATICAL COURSE. JUNIOR ALGEBRA. BY IT. y . lDTJI>TJrS, M". JL., PROFESSOR OF MATHEMATICS. Published by John Henderson & Co. 1882. LP tes. 9-^011 c. Ji > 1/a.j U r^- 9 ^ ^H- J ^^' ^ / /-f V F^ l^-^: PREFACE. The present work comprises the Junior Algebra of the course for the degree of B.A. in Queen's University. • Having been written for the use of students in class it is unlike the majority of works upon algebra offered to begin- ners. It contains no very elementary portions, as the stu- dents are supposed to be able to matriculate into the Univer- sity before taking up the work. It contains no lists of exercises, since it is expected that the Teacher will select or frame such exercises as may best suit his. immediate purpose. The work deals mostly with principles, and the examples, which are fully worked, are introduced for the purpose of exemplifying these principles. For the fuller elucidation of principles a Teacher is supposed to be available. In the establishing of theorems and formulae the method followed is inductive rather than analytical, as the former method is believed to be fully as satisfactory as the other, and much more within the grasp of beginners. The work consists of a small edition, and its production is somewhat of an experiment. If it is found to serve a useful purpose it may be followed by a similar work upon the Junior Geometry of the B.A. course. n. f. d. ERR A TA . << for 3 x — i The following corrections are necessary Page 12, line 5, 12, Ex. 19, line 5 16, line 12 47, Art. 52, line 4, 60, Ex. 103, line 6, 63, Ex. 106', line 3, 72, Ex. 115, line i, 92, line 3, " 118, line 2, " 2^ + 3r " =1 '* «2r « 2« ea< d 2 « x'z (( x + i \ q \ r)'^ = {p-\ q)^ \ ^i/i i q)'^,- ' .-. Mp\q){q\r){rip)'-^{/){q-\-r)^ p^ -q^-r^. v 2. Symmetry. — When an expression involves two or more letters in exactly the same way it is said to be symmetrical with respect to these letters. In writing such expressions we usually ignore the alphabetical order of these letters and give attention to their cyclic order. Thus, a'{h~c) + h^{c~a)+c\a-b) is symmetrical with respect to a, b and c, and in every term the letters follow the same order, a b c a b. a\b^c~d) + b\c + d~a) 4 c»((/-f « - b)-\-d'{a -I 6 - c) is symmetrical in a, 6, c and d ; but {a^-b+c■^-d? + {a^b-c-d)'^^{a■--b^c-'df ^{a-b-c + d? is symmetrical in 6, c, d, but not in a, since a is positive in every term, while the others are each positive in two terms and negative in two. The study of symmetrical expressions is of very great im- portance for many reasons. The principal one at present is that having the expansion of one term of such an expression the expansions of the remaining terms may be written down at sight. Ex. 6. To find the value of .s(s - 2a)(.s — 2b) + s(s — 2/; )(.s- -■ 2t) + .S(S - 2C)[S ~ 20) — (v — 2a)(s — 2b){s — 2c) when s=^a-\ b f c. This is evidently symmetrical in a, b and c. H{s — 2a){s — 2b) = !i{s'^- 2S'.a-f-6 + 4a6), .*. s(s-26)(s-2c)=:s(s* — 2S.6+C + 46C), s(s - 2c)(s — 20) =s{s^~ 2S.C + « + 4co) ; erv term the — 7 — and their sum is, si^.ab-^bc-^ca-s*). Hut (s- - 2rt)(.s - 26)(.s- 2t) =s* - 2s'(a + 6-hc) + 4s(rt64 /jc-l-cfl)— 8rt/)c = s(4.a6 t bc+ca — s^)~8abc. .'. the whole expression becomes Habc. ' 7. To sliow that 8(rt + fc-Hc)»-(a I 6)» -(6 + c)8 - u+a)* = 3(2af /; f c)(26-: f+4)(2r-l rt+6). In cases of this kind we may either bring one of these ex- pressions to the form of the other, or we may bring them both to the same third expression. The latter method is usually the simpler one, but the former is a better exercise of ingenuity. 8(a + b + c)^ = i2a + 2b+2c)^ = {a^b-{-V+c+cTa)^. .". Denoting a + b by />, 6 + f by q, and c-f a by r, we have (Art. i, Ex. 5) ip^q + r)^-p^~q^ r^ = ?i{p + q){q+r){r+p) = 3(rt + 26 + c){b + 2c + a){c -{-2a + b) = i(2a-\-b + c){2b + c + a){2c+a-\-b). 3. Multiplication by detached Coefficients. — In multiplying to- gether polynomials with one leading letter it is often advan- tageous to work upon the coefficients only, and to supply the leading letter after the completion of the work. Ex. 8. To multiply x^ + ^x"^ -2^+1 by 2x^ -x+2 I + 3 - 2 t I 2 -I + 2 2 + 6 -4 + 2 — I -3 + 2 — I + 2 + 6 -4 +2 8. Product : 2X^ + $x* - e^x^ -I- lox^ -^x-\-2 To multiply ^x^ —x-\- 2 by x^ + 2x'^ - 3. Here we must supply zeros for the coefficients of the missing powers of x, 3+0-1+2 1+2+0-3 &c., &c. 10. — 8 — To multiply ax^ +bx^ +cx-}-d by px"^ -\-qx+r. a + b + c + d ap +bp +cp -{-dp + aq +b(j +cq +dq -\-ar -\-br -\-cr -Vdr x-Ydr. ' apx^ + bplx* +cpx^+ dp aq\ bq ar x'^+dq cq br cr By observing the form which the product here assumes, and the manner in which its terms are made up, we may write it down at once in any similar case. Thus, 2ax^ + bx"^ — cx-\- 1 f jr^ ~bx + 2 2acx^ — 2ab x* I- 4a x^ + 2bx^ ~ 2c\x+ 2 be II. To multiply 2x+^y +z — i by x+y —2z + i. We readily see in this case that the product must contain the combinations of letters, x^, y^, z^, xy, yz, zx, x, y, z, and a numerical term. Hence we may arrange as follows : X* + 4a -62 x^ + 2b + be X^~2C. - b -C2 + c 2-3 + 1-1 x^ y' X y z 2 —3 1+1-2+1 z' xy yz zx X y z n -2-1+7-3 +1-4 + 3-1 product is 2x^~ ^\y'^ - 2z^ — xy-\- yyz - -^zx + ;r - 43; + 32 - 1 . 4. Multiplication of Series. — It often becomes necessary to square a series, or to multiply one series by another. In nearly all such cases the terms of the series are arranged according to the ascending powers of the leading letter. Multiplication of series finds its application in the algebraical development of functions, &c. Ex. 12. To multiply a + bx-^-cx"^ -{■dx^-\-. . : . by fl' + 6V+c';r2+rf';r8 + Product aa + ab' . ba x-\-ac' x^-^-ad' bb' be' ca cU da' x^ + . — 9 — By observing the mode of formation of the various co- j efficients we are enabled to perform such multiphcations with ' great facility. For example the coefficient of x^ is formed by taking the terms, a h c d J' ' L' ' a c a multiplying each pair together and taking the sum of the products. Ex. 13. To square the series i •¥ax-\-bx^ -\-cx^+ .... i+ax+bx^+cx^-{- .... Square = i + 2ax -j- 26 a' X^ +2C 2ab x^^ 14. To show that the square of the series j x^ x^ I -1-4:+ — +^+ . . • • 2 6 is formed by writing 2X in the place of x. i+x+^x^ +^x^+ .... square = i + 2;r + 2x'^ +%x^ + = i+(2X)+^-+ ^ 2 6 (2;tr)a + Ex. To multiply i +x+ 2x'^ -+-4x1 + . . .hyi-x-x^-x'^-. . . Ex. Show to three terms that if the series i -^x^ +^t^^ • • • and x-^x^ ... be squared and added the sum is unity. 11 lO DIVISION. 5. Division, being the reverse of multiplication, may like that process be carried out upon the coefficients, and when only one letter is involved in the expressions under considera- tion, the process becomes in this way very much simplified. If we multiply Ax^ -j-Bx+C by ax^ +bx-\-c, we obtain for coefficients, aA -\-aB hA i-aC bB cA + bC cB + fC By observing how the terms in this product are formed we may reverse the process and thus perform division. Thus, if a + b+c be the coefficients of the divisor, we see that a A divided by a gives A , the first term of the quotient : then bA subtracted from the second term and the remainder divided by a gives By the second term of the quotient ; and lastly, cA +bB subtracted from the third term and the remainder divided by a gives C, the third term of the quotient. In this operation the only quantity by which we really divide is a, and hence if this be unity its presence may be quite ignored. Again, since algebraical subtraction is equivalent to addition with a changed sign, we make our subtractions additions by changing the signs of every term, except the first, in the divisor. Ex. 15. To divide ^x^ - ^x^ — ^x^-^yx - 2 by 2x^ - ^x-{- 1. We may write the divisor in any position which is convenient. as, 2 + 3 -I 4-4-5+7-2 +6+3-6+2 - 2 - 1 ' or. 2 + 3-1 4-4-5+7-2 +6+3 -6+2 2+1 -200 -2-1 2 + I' — II and the quotient is 2x'^ The second position oft: most convenient. V-2. } divisor is for some reasons the Ex. i6. To Divide Jtr® — 3:r^ -f 6jr- 4 by 4r* — 24r + 1. Here the coefficient of x^ of the divisor being i we may ignore it altogether. 2-1 i + o +0 i-o — 3 + + 0+0 + 6-4 24684 0—4-8+4 —1—2—3-4—2 0+2 1+2+3+4+2 0-2-4 o O .*. quotient = x"^ ^ 2X^ + yir^ + 4X* + 2x^ - 2X - 4. 6. If the case is one of inexact division, we must stop the process at a certain point if we wish to obtain the correct re- mainder. This point is of course reached when the last obtained term of the quotient does not contain ^• In order to determine this point we draw a vertical line to the left of the divisor as usually written, i. e. between the first and second terms of the divisor as completely written ; all the terms of the quotient proper are to the left of this line, and no term of the quotient line to the right of the vertical is to be used in forming a partial product in getting the remainder. Ex. 17. Divide ;r' -x^ -\- ^x^ + io;r'' -5:1^-1 by x* - 2X^ +x^ -2. 2- 1+0+2 i+o-i+0]+5+ 10-5-1 244: 4-244 -I -2i -2+4 I 2 1 + 2 + 2 + 2| +9+ 12 -1+3 .*. quotient = x^-j- 2X^ + 2X+ 2, and remainder = gx^ -\-i2x'^ —x-\-^. 7. In cases of exact division the process may be carried out in a somewhat similar manner when several letters are in- valued. Ex. 18. To divide />" +pq + 2pr - 2^' + yqr - ^r' by p - q -j- 3^. — 12 Since the dividend is of two dimensions and the divisor of one, the quotient must be of one. We may arrange as fol- lows : P"" q' r" pq pr qr p - 1 1-2-3 I 3 7 9 ' I -I -2 -7 r -3 P q quotient = p-\-2q-r. Here the coefficient of ^ in the quotient must be i so that when multiplied by that of^ in the divisor it may cancel that of/>2 in the dividend. Similarly we find that of q and of r. Then for Pq we have -1.2 + 1.1= - i which cancels that of pq in the dividend, &c. Ex. 19. Divide 2!^ - ^x^y - ^x^z - ^xy^ - j.r-j* + 1 2xyz h 2y^ - zyh — 2,y^ + 2^ by x->ry - 2Z. The quotient may obviously contain all possible terms of two dimensions, and can contain no others. X^ y3 x'^y x^2 xyz y^z xy"^ xz'^ yz'' X y — I — I 2 2 -3 -3 -2 4 12 -3 -3 -I 4-2 - I -I 5 — 10 3 -3 1 I 2 2 1-5 I I xz yz . . . . possible terms of two dimensions. j/2 ^2 xy quotient 2x'^ + 2y'^ z^ - ^xy-^xz-\-yz. 11- 1; ■ 8. In a case of inexact division, as in example 17, if we neglect the vertical line and its indications the quotient will extend to an indefinite number of terms, which will follow a certain law of formation, and it will thus become an infinite series. This is very similar to cases of inexact division in arithmetic when the quotient is run out into a circulating decimal. g. Expansion by Division. — Let it be required to divide i by i-^x, running the quotient into a series ; we obtain, terms of ■Z'i ysl 3 -3 I I 2 2 • • terms of 13 I i+x = 1 -x + x^-x^ + x*- + . . . . 1 1 similarly = 1- X = i+x + x^ + xr' + jii^++ .... By means of these forms we may effect the expansion of any expression which can be expanded by mere division. Ex. 20. To expand -; into an infinite series. b + x a _ a b+x b ' ^ =«fi- ^+^-^V-.. ..V • X bV b^ b' b'^ b 7 Ex. 21. To expand 20^- 1 "'-i-d- I -2a^ +. . . .). 20^-1 • •/> = 1+0 [2 +2(1* +4a^ + . . . . Ex. 22. To expand (i-.-y^ I I +z {i-zf 2 -^2 - r:-(l +Z)(l + 2Z — Z^ -^ 2Z - •f-. . .) — 2Z + Z^ = 1+3-- ^5z^ + 7^ + Ex. 23. To expand .x"^ - 2;i; - I I+X + X^ ' This becomes (, X^-2X~\){l-X + X^ + X + }? -. . . . .), = -I- X+IX'^-2X'^-X*+IX^ .... yjitr^^ €^y / f fiUr^' ; V, — 14 — SUBSTITUTION. I s .:,1| li :vi 10. Substitution is the writing of one quantity for another in an expression. Thus, Aa^ + Ba^ + Ca-¥ D is obtained from Ax^ + Bx^ +Cx +D, by substituting a for x. Let A' stand for a general expression of the form Ax^' + Bx^'-^i- Sx+T, where n is an integer greater than unity. If we divide this by ;r - a we will obtain another expression one dimension lower, which we may denote by A'l and, in the case of inexact division, a remainder R which does not con- tain X. Hence we may write, Ax'' + Bx''-^ + . . . . Sx+r=:X, (x-a)+R. Then, i. If we substitute a for x throughout, we get A rt" +Ba"-^ + . . . . Sa + T = R. Hence we conclude that if we divide an expression contain- ing only positive integral powers of x by x~a the remainder will be the original expression with a substituted for x. ii. If x — a divides A' exactly, R is nothing ; and substituting a for X we have, /4flM-B«"-' + Sa-{T = o. Hence if x -a is An exact divisor of an expression contain- ing only positive integral powers of x, the substitution of a for X in the expression causes it to become zero ; and conversely, if this substitution renders the expression zero it is* exactly divisible by x - a. — 15 — II. Applications of i. and ii. I Ex. 24. To find the remainder when ;r'-3;ir*4-2 is divided by ;r-i. Remainder = i' -3.1^ + 2=0. / ^ .*. ;t' - 3;t* + 2 is exactly divisible by x-i. In a large number of cases the substitution is most readily effected by means of the division itself. E\. 25. To find the remainder when p^ - 2p* +3/> - 10 is di- vided by /) - 4. 4 I -20 o 3 - 10 ___ 4___8 35 128^ 524 2 8 32 131 514 = Remainder. Hence to substitute a for ;«; in A' divide the expression A' by X - a and take the remainder. Ex. 26. To find the value of n'* - yi^ +2n-^ 10 when -3 is substituted for n. Divide by n + 3. -3 10 -266 I 00-3 2 -3_ 9 _-27_„90 -39 - 30 92 -266 .". Result = -266. Ex. 27. Is x—2 a divisor of -t* — ^x^ — ^x"^ +2;r+ 20 ? If we substitute 2 for x we obtain 2* - 3.2^ - 4.2'' + 2.2 + 20= o, .•. X--2 is a divisor. Ex. 28. Is fl — 6 a divisor of ab(b -a) + bc(c -b)+ ca(a - c) ? Substitute b for a, and we obtain, fe" .0 + 6c(c - 6) + c6(6 - c) = o, .'. rt -6 is a divisor. 12. In transforming equations it frequently becomes neces- sary to substitute a binomial expression for x in the general expression Ax'^ + Bx^'^ -\-. . . . Sx-^T. This may obviously be done by writing the binomial in the place of x, and then ex- panding, as follows : ii; - i6 - « Ex. 29. To substitute jy - i for x in x^ - 3^' -|- 2,r -t- 1. We have, {y~i)^ - jO' - i)" 1 2(y- i)-i-i, =^'J - 6)'^ -f I ij* - 5, by expansion. The following will enable us to perform this important sub- stitution more readily : ^ Since x—y- i, :.y = xAr\, and the given expression is to be put under the form (^-f-i)3-f/v'2(4r+i)2 4-/e,(;r4i)+/e; Where we have to determime the remainders A*, /?,, A'j. If we divide the original expression by x-\-\ the remainder is A'. If we now set aside this remainder and divide what is left by ;r- I the remainder is A'l. Proceeding in this way we obtain all the remainders. The whole operation is as follows : -3 -I -4 — I -5 - 1 -6 4 2 4 6 A II -I + 1 -6 .*. (,r-|-i)^ — 6(;lr^ i)^ -|- ii(;tr-f i) - 5 is the expression : or, >'3~6)'2.| 11;;- 5. Ex. 30. Express 3/)^ ~p^ -V 4p'^ -\- 5/) - cS in terms oi p - 2. 6 - 1 12 4 22 5 -«L2 52 114 6 6 II 24 26 70 57 192 loC 12 6 35 36 96 142 249 18 6 71 48 238 1 24 6 119 30 — 17 — I 249(^-2) + 106, is tho ref|uirefl expression. Ex. 30'. In 111^ --^m'^n \ imn"^ jm^ substitute m - n for w. .J 2 -.J I - 2 0, -2 -J I - I - 1 I I .'. (m - «)•'-«*(;;/ -«) - 3«'* is the expression. 13. The following form of substitution is of importance in many operations. Ex. 31. What does x* - ^j^ +2^-1 become wlien x^ 1 ^- i - o ? This may be solver! by division directly as follows : - I I I 4 2 - I - 1 I -2 3 - I 4 - 4 I - I 2 -3 I I - I 2 _____ I "4 7 -5 .•. 7^-5 is the result. Or it may be done thus : •.* x'^ -\rX -1 = 0, .*. x'- X* = \ \ x'^ - 2X~ 2 S^' X'* -X rl^ 2X - I and x"^ --= x*.x^ - (2 - ^x){2X - i ) = ij.r 8. .-. ;r' - ^x^ 1 2x~ I -yx - 5. 14. We will now extend Art. 1 1 to the case where the num- ber to be substituted is partly a whole number and partly a decimal. Ex. 31'. Find the value of x'^ - ^t^^ -\ 2X + i, when ;r:= 2.85. In this case we work through for each figure separately, as follows : — i8 — I I:'. I. 11. in. case -3 2 4-2 - 2 + 1 1 2^85 the an e -I 2 I a 2.. I . . . Ex. 2 3 . .8 3.04 4.032 3.8 8 504 3.68 5032 . . . 4.6 8 5-4- 05 545 8.72 . . .2725 8.9925 •449625 5.481625, result. Kx The above work is fully expanded in order to show the various steps. We first work through for 2, as in fornner ex- amples. Then we work through for the 8, remembering that as It is in the tenths place the figures in column I. will be moved one place to the right, two places in column II., and three places in column III. And in like manner for 5. The work may be very much condensed, as follows : Ex. 32. Find the value of y* - 4_y'* - I wher i;'=2.i3. I - 4 •I 2.13 2 4 8 6 20 81 2081 82 2163 83 2246 843 227129 16 1 808 1 20244 20925387 ■I .8081 1.43586161, result. 15. If we have an expression j*^ -rt and we put for jf any particular value we have seen (Art. 14) how to find the value of the expression. If that value is zero, then y^-a = o, and therefore y = ^^a. Hence, if we can discover a quantity which when put for^* makes the expression y - rt zero, that quantity is a cube root of a. And similar reasoning would apply in the 1.85 — 19 — case of any other root. The process of the last article supplies the means of approximating to this value and thus becomes an elegant means for the arithmetical extraction of roots. Ex., 33. To find the cube root of 2299968. result. show the )rmer ex- iling that '. will be II., and 5- I, result. ■ Jf any e value = o, and y which juantity V in the - 2299968 I I + 1 2 33 36 392 3 399 507 51484 -1299 + 1197 - 102968 + 102968 extract the cube root of j. -3 .1^ X I I 2 34 38 424 428 3 436 588 60496 62208 -2000 1744 - 0256000 0241984 ..14016 132 44 We know that i is the first figure of the root ; we, therefore, work through for one. We then find the next figure of the root by employing 4 in the second column as a trial divisor, and 20 in the third column as a dividend ; but as the 4 will be increased by the subsequent operation we make a proper allowance in the quotient figure. The principal points to be attended to are that the number carried to the third column must always be less than the number above it from which it is to be subtracted, and that the remainder after subtraction must not be greater than the last completed number in the second column.* After obtaining 3 or 4 figures the number of figures may be doubled by employing the last completed number in the second column as a divisor and the last remainder in the third as a dividend. Thus dividing 140 16 by 6221 we obtain 225 ; hence the cube root of 3 is 1.44225 true to the last figure. In a pre- cisely similar manner we may extract fifth and seventh roots, &c. * In some special questions it may be greater by a small amount. 20 — OF FACTORS AND FACTORING. x6. In the expression a(6 -c)( -|-c is a trinomial factor. These are linear factors, containing terms of only one dimension, while a'+2a - i and ab + be are quadratic factors, inasmuch as they contain terms of two dimensions. An expression may have real quadratic factors when it has no real linear factors, e. g. ** + 3^' + 2 = (X^ + l){X* + 2) in which neither of the quadratic factors has any real linear factor. 17. Theoretically any expression of the form Ax'' + Bx''-^ + Sx-^-T may be written as the product of n linear factors containing X, as A{x-a){x-fi){x-r) .... U-c), in which the values of a, /9, &c., depend upon those of A, B, C, &c. ; but practically the discovery of the values of «, ^, y, &c., cannot always be effected by any means at our command, so that the actual process of factoring can be carried out only in special cases. These are, however, frequently of great importance. Only the simpler processes of factoring, will be dealt with here. 18. Factoring by reference to known formulas. The formulae more generally useful are : i. a'-b''={a]-b)(a-b), ii. a» + 62±2«6 = (a±6)^ — 21 — iii. x'*-ir(a-\b)x-ial) = {aix)(l)-^x). iv. rt' t 6' + c' -f 2(ab -f ic -t c «) = (« + 6 -+- c )', V. a^ + b^ + c^-2ahc = {a-ib-\c){a.a-h\h^b'-c-\-c.~a). vi. (rt i b -}- c)' - rt' - ^^ - 1» = 3(rt i '')(/; i- c){c f rt). ICx. .55. X^ - 2X - ^= X^ - 2X + I - 4 = X - I - 2^ -^ {X - ^i)(X f- I ). or .t^-24r-3 = 4r'^ (I -.i)4r-| I X - 3 = (;r| i)(:r- 3). Kx. 36. m'' + 4;« -6=m' + 4m + 4- 10 = (;;;-[ 2)'- | lo'', = {m + 2 1- |/ io)(;;j + 2 - | 10). E\. 37. rt'^ 4- 2ab -i b'*~a -b-6= {a + bf - (a \ b) - 6 = (fl + f>)' - {a + /^) -I i - V = (" + -"^ - .JK)+a(rt + 2^)-| 2^( iil i .: i m a{b-vbc -c) -vb{(:Arca-a)-\-c{a-^ah ~b) - n.abj * must be true quite independently of any values which we may give to a, b and c, that is it must be identically true. Make then, a = I, ^ = 2, c=: 3, and we obtain i8=6», .". « = 3. Hence, ^ a{b + bc -c) -\- b{c -i-ca-a) + c[a + ab ~b)= ;^abc. y Ex.40. To (sictor ab{a—b)+bc{b — c)-^ca(c -a). Putting a = 0, we find no monomial factors. Putting a-b = 0, or b = a we find a—b, and hence from symmetry b — c and c- a to be factors. .'. the expression is equivalent to ii.{a -b) {b-c){c-a), and we readily find n= - i. .". ab{a — b)+bc{b—c}-^ca{c- a) = -{a-b){b -c){c-a). Y Ex. 4 1 . To factor 2ac{2a - c) \ 2cb{2c - ^) + 2ba(2b -a)- jabc. We readily find this to be equivalent to - {2a - c){2c - b){2b - a) . f" Ex. 42. To factor ab{ll'-a}) -\- bcic" - b"") +ca{a'-c''). This is symmetrical in a, b, c, and is of four dimensions ; hence there are four literal factors. We readily find that a—b, b — c, c — a are factors. And since the expression is symmetrical in a, b, c, and can contain only one more factor, it also must be symmetrical in a, b, c. There- fore it can only be a + ^ + c. And the expression is equal to {a - b){b—c){c — a){a + b + c). V Ex. 43. ToidiCtov ab{c—d)-\-bc{d — a) +cd{a — b)-\-dii{b -c). Putting a — 6 = o, or 6 = a, a'^{c-d)+ac{d—a)+da{a—c)~o .'. «-6isa factor, and from symmetry 6 — c,c-^, and -l^^cJf^ + P. a^ -ad^ + a'^6 -- b^ = {a -'r b){a -\ b){a - b), a^ + ^a'U + :iab'^ + b^ - in -r b)(a ^ b)[a rb), .-. H.C.M. = [a+b]'^. 22. If two expressions na, nb have a common factor n, their sum, their difference and the sum and the difference of any multiples of the expressions will have the samec(jmm()n factor. For, na ± nb = n{a ± b) ; and na.p±nb.q = n{ap ±bq). This lies at the basis of the common method of finding the H.C.M. of given expressions. Ex. 47. To find the H.C.M. of 6.y3 2.v:^ + 3-^'^ -ii;r- 6. Taking coefficients only, a ... 6 - 7- 9-. 2 3/9. . . 6-j- 9-33-18 3/9-a . . . 16-24-16 Divide by 8, 2-3-2 .-. 2.^2 -3;r- 2 = H.C.M. yx"^ - ox - 2 and ^ . . . 2H-3-II— 6 3a ... 18 -21 - 27 -6 3« -/9 . . . 16 — 24-16 -i-Sx . . . 2-3-2 r. — 25 — Ex. 48. H.C.M. of6;r8 + i5;r''»-6;t + 9and9;r8 + 6;r2-5i+36.: Here we can divide through by 3, which will be a factor of the H.C.M. a. . . 2+ 5- 2+3 3a. .. 6 + 15- 6+ 9 2^. . . 6 + 4-34 + 24 r 11 + 28-15 5r- • -55+140-75 II5. . . 55 + 198 + 99 58+174 -58 1+3 .-. H.C.M. = 3(4: + 3). ^. . . 3+ 2-17 + 12 4a. . . 8 + 20- 8 + 12 diff. -i-x = d 5 + i8+ 9 5^. . . 25+90 + 45 3r- • . 33+84-45 58 + 174 58* 1 + 3 Ex. 49. H.C.M. of ioy^-\-y^ -gy + 24 and 20_y* -17^* + 483' -3 a. . . 10 + I — 9 + 24 5^. . .10-5 + + 15 e 6-9+9 88-a 6-9 + 9 j3. . . 20 + 0-17 + 48 — 3 2ay. . . 20 + 2-18 + 48 d 2- 1+ + 3 .•. H.C.M. = 2j'''' -3>' + 3. LEAST COMMON MULTIPLE. 23. The least number of which two given expressions are factors is their least common multiple. If flcg/ and a^^^ be two expressions, their L.C.M. is adcdef since this is the lowest expression which contains both. To find the L.C.M. of two quantities we take the factors which are common to both and those which are peculiar to each and multiply them together. Thus a, e, are common to both the foregoing expressions, c, /, are peculiar to one, and if, d, to the other. Ex. 50. L.C.M. oi x^~{a-b)x-ab and x^ -2ax-\-a^. By factoring these become {x—a){x + d) and (x—af, ^ - ^ ^^ 26 .'. (x—a) is common to both, {x + d) is peculiar to the first, and the second (;r- a)to to the second. .-. l^.CM, = {x-af(x-+lf). 24. If there be more than two quantities we proceed in a T a a_ _ ax + by a \ !il|r- 28 Ex. 56. -h + x+jf (x-y){x-z) {y-x){y-z) {z-x)(2-y) By putting in one factor in each denominator and arranging in cyclic order we have, -yiy - z) +■ - x{s — x) + - -{x+y){x-y) {x-y){y-z){z-x) {x - y){y - z){z - x) {x-y){y-z){z-x) _ ziy-x) _ z _ z Ex. 57. {x-y){y-z){z-x) {y-z){z-x) (y-z){x-z)' a . b . c + + {c-a)(a-b) {a-b){b-c) {b-c){c-a) _ a{b -c)+ b{c - a) 4- c{a -b) _ Ex. 58. (a-6)(6-c)(c-fl) since the numerator is zero identically. ,i a b + (a-b){b~c){x-a) {b-a)ib-c){x-b) {c-a){c -b){x -c) The common denominator is (a -b){b- c){c - a){x - a){x - b){x - c) . The first numerator = - a{b -c){x- b){x - c) and by symmetry the others are j ~-b{c-a){x-c){x- a) { -c{a-'b){x-a){x—h) .'. whole numerator - - \a{b - c){x-b){x-c)-\rb{c-a){x-c) {x~a) + c{a-b){x-a){x-b)\. Now a - 6 is a factor of this, and therefore b-c and c - a are factors. To find the fourth factor which probably contains x, let the factors be, {m + nx){a - b){b -c){c~ a) where n is numerical and m may be so. To find them put a = 2, 6 = 1, c =0, and we obtain, m = o, n= 1, .'. the factors are, x{a - b){b -c){c -a), which reduces the whole fraction to X {x-a){x-b){x-c). Ex. 59. 6c- + — 29 — cd + da (a~b){b-c){c-d) (b'C){c -d){d-a) (c -d){d-a){a -b) + _«6 ; {d-a){a-b){b-c)' The numerator in this case becomes, ab{c~d) + bc{d-a)+cd{a-b)+da{b-c). a- 6 is a factor of this, and from symmetry b—c, c- d, and d -a are factors. But it cannot have four literal factors, therefore it must be zero identically ; hence the sum of the factions is zero. 27. The following relations among the terms of fractions can often be employed with great advantage in algebraical trans? formations. They are useful in reducing fractional expres- sions, and they lie at the basis of the relations employed in proportion. I. If a T 1. — = — -, for a:!^=.^=i.toshowthat cr (r c^ {X ^y + ^)(aa -f ** + c») « a^x ^if^y-\- c^z Multiply the numerator and denominator of the first fraction by X, of the second hy y and of the third by s. Then summing numerators and denominators, 3^-\-^-\-x^-ycyz a^x^-lf'y^ch = I or factoring, .^yi^^'^^l-^ -yz-zx) ^ ^^ a^x f Iry -1- rz l^ut from the original fractions, x^-'ry^'V^ - xy -yz-zx=a^^-lf^^\■c^^ RATIO, 28. Th€ ratio of one quantity to another is the numerical quotient which arises from dividing the one quantity by the other, or it is the number which expresses how often the one quantity is contained in the other. Hence a ratio is an ab- stract quantity, and in order that magnitudes may have a ratio the one to the other, they must be of the same kind. Thus there can be no ratio between miles and years although there is between numbers expressing aggregates of miles and years. If a, b, c, denote certain lengths, a has a certain ratio to b and to c, but a has no ratio to be, since be denotes an area. And thus in geometrical applications of Algebra the terms of a ratio must be homogeneous. But if a, b, c, denote numbers, any combination of them may be employed as terms of a ratio. The ratio of a to 6 may be expressed either as — , or a:b. In any case a is the antecedent and b the consequent. If a is greater than 6 it is a ratio of greater inequality, if less than b of less inequality, and if equal, a ratio of equality. The ratio b'.a — 3^ — is the inverse ratio o( a.b. The ratio fl'rft' is the duplicate ratio o(a:b; fl':6'the triplicate, and rt*:/>^ is sometimes called the sesquiplicate ratio of a:b. The doctrine of ratio is extremely important in modern mathematics, for it frequently happens that the terms of a ratio are of little or no importance while the ratio itself is all-important. We have examples of ratio in the circular measure of angles, in sines, cosines, tangents, specific weight, &c. The propositions of Art. 27 apply directly to ratios as frac- tions. From these it is evident that a ratio is not changed when both terms are multiplied or divided by the same quantity. 29. Let rt:/; be a given ratio ; then dividing both terms ')V X, — : — is the same rs a:d ; but when x becomes infinitely XX great each term becomes infinitely small. Hence quantities which become infinitely small, and are thence called vanishing quantities, may have a definite and finite ratio. This princi- ple lies at the foundation of the Differential Calculus. Ex. 63. What is the ratio of {a -\- x)"^ - a"^ to x when x becomes infinitely small ? ratio = i.a-'rX)'^.- C^-2GX^-^ 2ax-{- x^ — 2a-\-x=2a when x becomes infinitely small. Ex. 64. To find the ratio of s;!:"- 24: +2 to x^-\-x-i when x becomes infinitely great. Ratio = 2 , 2 3- +. 3x'-24r+2 ^ jp j^ = ^ = Z- = 3 when 4: = 00 . X*+X-l X x^ 30. Prop. The addition of the same quantity to both terms of a ratio of inequality brings it nearer to a ratio of equality. — 33 — Let — be the ratio, which suppose greater than i. then-r-- I = — ; — 6 b Now add x to each side, and, ? — - - i =?-^ . b + x b-\-x Kut 5 is less than — j—, . . -, b-t- X b b-\-x is nearer unity than-^is. b A similar proof applies when a\b is less than i. 31. Ratios are compottnded by taking the product of the antecedents for a new antecedent, and the product of the consequent for a new consequent. PROPORTION. 32. When two ratios are equal the terms taken in order are said to be in proportion, or to form a proportion. Thus, . if _- = __. , then a, b, c, d are the consecutive terms of d a proportion, which is often expressed as a'.b'.'.c'.d. a and d are the extremes, and b and c the means. The terms a, b, as also c, d, constitute a couplet ; and the proportion is read, a is to 6 as c is to d. If the terms of the last couplet be divided by c d, we have, 1... t . I a.b .. -J- . — , d c or a is to 6 inversely as d is to c. 33. The following variations in a given proportion are di- rectly derived from Art. 27 ; some of them have been distin- guished by special names. — 34 — ifA/iO^'*^ If a'. h::c:d, 1. a.c.:b:d, . . . . Alteinando. 2. b.a::d\c, .... Invertcntlo. 3. a + b.b.:c + ci'd,.... C'onJiionunilo. 4. a — b'.b\\c-d\d,.... Dividondo. 5. a\a-\b\\c.c id Convintfiulo. 6. a^b\a-b\\c^d\c-d,9^^^-^A]^i-^'y^'^^ If«:^::cur.:f:/, 9. a\b::a-\-c-\-e\b-¥d-\-f. 10. fl : ^ ■■ : ma + nc -j-pe : m^ -f Wf/ +/»/. 11. rt":<^"::a" + c"-f<'":^" + ^"+/". U a.b:\b.c..c.d, 12. a: cwa^.b"^. 13. a:', 6 or, 2". 2"+^ ^=2". 2", 6 .-. ;;4- I = - . and n evidentlv is 2. 40. An expression denoting;- a root which canriot be exactly obtained, as | 2, 1^5, ike, is called a surd, or irvaiional quan- tity. Surds are divided into orders dependinj^ upon the index of the root to be obtained : if it be a square root wu have a quadratic surd, if a cube root a cubic surd, ik.c. , The product of a rational quantits with a surd is known as a mixed surd, but when all the factors are under the surd sign it is termed an entire surd. 1 1. Surds may be indicated by indices as 2"^, 5"', = (<'fl^6. IV. y a.f a=^u^ .a^ - a^ '^ —a**. V. , p^q^} p'^.y q=p] q. l^x- 75- I 3-1 2.V 5 = 1 30. Ex.76. I 184 = 1 2^.46 = 21 46. Ex.77. ':-^:^^-^ = ^ 3. 31 2 31 2 3 Ex. 78. I^54rt*-if2j,3 _ f' 2ya^y^.2ax^ =^ay^2ax^. 42. Fractional expressions with a compound surd in the denominator are simplified by rendering the denominator rational. The methods of doing this are shown in the follow- ing examples : Ex. 83 Ex. 84 Ex. 85 — X 2 - 1/2 2-1/2 2 + 12 •IZii-? = (3.-y2Mj. 2 + 1) ^ J y' 2-1 l±J^ = i±J_2.i^^,,2. + 2l/2. 2-1 2 _ 41-1^2 + i/3)_ I -\ 1/2-1/3 21/6 - 4 _ 2{ I - |/ 2 + J/ 3j(2 V 6 H- 4) ^ (l -y 2 + v 3)(t/64-2) 24-16 2 _ 1/ 6 + v'2 + 2 43. The following propositions with respect to quadratic surds in particular are important : i. The product of dissimilar quadratic surds cannot be rational. For, let I /> and 1 (/ be their simplest surd factors; then neither /> nor q contains square factors, and being dissimilar they are not made up of the same factors ; therefore, their product cannot be made up of square factors, and consequent- ly \/pq is not rational. ii; — 42 — ii. A surd cannot be made up by combining rational quantities and surds by addition and subtraction. For if possible, let y'p - m ± \,^n ; squaring, p = m^ +n± zm \/n ; .'. |/»= ±C-— " = a rational quantity. 2m iii. A surd cannot be made up by combinini^ two dissimilar surds by addition and subtraction. For if possible let V/» = Vq± \/y squaring, p=q-\-r±2 \/qr, .'. ^qr = ±^{p -q -r)= a rational quantitity. But since q and r are dissimilar, \/qr cannot be rational. iv. If x+ ^/y = a-\- 4/6, then x - ^'y = a - \/b. For X- a=: yb - i/y... But since ^ - a is rational it cannot be equal to the difference between two surds. Hence x — a=o, and \/''b-\/y=:o ; .*. x=a diXid \b = \/y ; and X - [/y=ci- \/b. 44. To find the square root of a binomial quadratic surd. Let Va + \^b = y/'x -f |/j. squaring, a + y/b =x-j-y + 2 y/xy ; .*. Art. 43, iv., x-{-y = a, and 4xy = b. Hence (x +y)^ -b = {x -y)"^ -a"^ -b, and .•. x—y= V a^ -b. V>wt x-y = a, ^ .-. x = \{a + Va'^-b), 2Sid y=^{a-\/a^ -b). Ex. 86. To find the square root of 3 + 2^/2. Here ;r+>' = 3, and ;ry = 2. 43 — and x-yzz\. Hfence, x-= 2, and^ = i, •"• 1^3+21/2= I +]/2. • . Ex. 87. To find the square root of 23-41/15. ^+^' = 23, and 4;ry = 240 ; .-. ;r -j/=i/(232 - 240) = 17 ; .-. ;«r=20,3'-3, and v/(23 - 4v'i5) = VS + 1/3- 45. In the case of trinomial quadratic surds which are com- plete squares we may proceed as follows : Let |//> + \/q + i/r be the root. Then (|//» + yq + \/y)'^ z=p + q + r + 2\/pq + 2\/qr + 2\/rp. But p = ^J^^Mlly^. 2X2y'qr Hence if P, Q, R, denote the surd terms, taken in order, P PR ,.QP ,^RQ Ex. 88. To find the square root of 54 — 41/2 +61/5-121/10. 2x6|/5 41/2.61/5 ^ . / 2X41/2 .*. I ±21/2 ±31/5, form the terms in the root, and a little inspection shows us that the signs must be 1-21/2+31/5. In cases of this kind the subsequent squaring of the root is the only sure test of correctness. 44 SURD HQUATIONS SOLVED AS LIN EARS. 46. Equation? containing surds can sometimes be solved as linears, but in all cases they involve certain peculiarities which will be more fully comprehended hereafter. Ex. 89. Given I a^+x^ + v'a^ -x^ = b to find x. Squaring, 2a^ + 2l^a* —x*= b'^ : transj.osing and squaring, 4a* - 4-r* = (6' - 2a*)^ ; .'. x=t {a*-\{b'-2ay\' Since the fourth root of a quantity has four values, x ihas four \alues which will satisfy the equation ; and thus the e(iiiatioi) although apparently solved as a linear, is in fact a quartic. Ex.90. Given ^a-^i/x=-i/ax to ihnd x. Here we reduce the number of surds containing x by divid- ing l.y I X, and and a (|/a — i)^' a X = (l/«-i)^' ^ ^. \/a + x , V'a + x i/'x , c j Ex. 91. Given, — - -f 1_ =!L. to find a x b CLXfT-' b .'. b[,a-\-x)^ = ax^. Squaring and extracting cube root, A 2 whence, x—- a ab'^ 1 2" 45 -- • OF THE QUADRATIC r'^UATION. 47. A quadratic equation contains the second power of the unknown quantity. If it contains that power only it is called a pure quadratic, but if it contains the first power also it is a mixed or adfected quadratic. This distinction is, however, of little importance. 48. Origin uf a Quadratic. If two linear equations, with the same unknown quantity, be multiplied together the product is a quadratic. Thus, (x-a=^o)(x- b=o) gives x'^ - {a-\-b)x^-ab=o. And conversely, every quadratic can be formulated as the product of two linears. Thus, '\{ x"^ +px + g=o denote any quadratic, ■-i^r _i_r» — " — n ^-«^>-# — — — :t' +px 4 4 = 0, TO .-. [X+±)^ -l(PL-gy2=.o, X + H*^T-^'}•.|--^^l<^-^)|=<'• In which the quantities within the | i are linear equations. Hence every quadratic may be considered as the product of two linears. 49. Roots of a Quadraiic. A quantity which, when put for the unknown quantity in an equation, satisfies it, or renders it true, is called a root of the equation. A linear has but one root ; but a quadratic, being the pro- duct of two linears, is satisfied by the root of each linear ; every quadratic has accordingly two roots.* Thus, if x^ -\-px + g = (x- a){x — b) =0, where />= -(a + ^) and g = ab, then;ir = a, or x-b satisfies the equation since either substitution renders the expression zero, hence a and b are the roots of the quadratic. * This Btatemeiit 1b not \«^ithout oxct i lionp, to some of yflilch reference will be made hereafter. -46 - 5o. General solution of a Quadratic. The most general form in which a quadratic can be written is ax^ +6x+c = o ; when a, b, c, denote any quantities whatever, and this we are to resolve into linear factors. Multiply throughout by 4a, and add and subtract 6^, and we obtain 4a*;r* + 4adx + 6^-6"^ + ^ac = o. .'. (2ax-hd)^ -{d^ -4ac)=o, or, {2ax+d-\-Vd^-'4ac)i2ax+d- i/d^ -4ac) = o. Whence if x^, x^, denote the two roots, ^1 = x^ = 2a -b-\-\/W-~^c {A). 2a These may be combined in one formula by using the double sign ± , and we get, _ -b±Vb'^ -^ac 2a (B). A study of this form serves for the solution of all quadratics. Ex. 92. Let ^x^ -2;r-f4 =0 ; then,;r = ^^^f-lg- = ^^<"^, = i(i±v/-ii}. Ex. 93. Let {a - b)x^ ■\-ax-\-b = o. then,;r=--^±^^^'-4fe(^-^). 2{a-rb) _ —a±{a-2b) 2{a - b) - — , or -I. b — a - 47 -• 51. Stun and product of ! he roots. Adding the values of the roots in Art. 50, (A), we obtain, and nuiltiplying, we obtain, c a X\*i - Hence in a quadratic the sum of the roots is equal to the quotient arising from dividing the coefficient of x by that of x^ taken with a changed sign ; and the product of the roots is equal to the quotient arising from dividing the constant term by the coefficient of ;r*. Ex.94. Given (s/+^)'' + (cZ + a)" = y*, and s'*+c'''=i, a qua- dratic in / to find the sum and product of the roots. Squaring and arranging in powers of/, .-. /j +t.. = - 2(c« + s/9), and /i/2 = aa+/92 ^r2. 52. Nature of the roots. In the formula Art. 50 (B), since 6* is essentially positive, and since a may be rendered positive by change of signs, the character of the quantity under the surd will depe.nd upon the «ign and value of c, a being positive i. If c is negative, then b^ - ^ac is positive, and has a square root either rational or irrational. Hence in this case the roots are always real quantities. Thus, \{ x^ i-4x — n = o, x has two real values for every posi- tive value of w. ii. Iff is positive and less than — , the quantity 6'- 4^0 is 4^ positive, and the roots are real. iii. If c, being positive and less than — , gradually increases 4a in value, then the two values of at, i.e. the two roots become more and more nearly equal as \? - ^ac becomes smaller ; and finally the roots meet and become equal in value when h^ — 4ac -48- becomes zero. The surd part then disappears and the qiin- dratic has equal roots which may be botli positive or both nejjfative, but which are always rational. The very important condition then that a quadratic may have equal roots is that ^ac = 6". Ex. 95. If r^- 2drc-^(P -a^^o be ix quadratic in r, find the condition that r may have two equal values. Condition, 4((f - «") = 4^/^, or, iv. If c be positive and greater than — , b^ - ^ac is negative, 4a and as the square root of a negative quantity has no real ex- istence but is wholly imaginary, the roots of the equation will be imaginary or impossible. These imaginary results are not to be dismissed as of no consequence, as they are frequently of very great importance. Let it be required for example to divide 10 into parts such that their product may be 30. If x be one part, 10 -;« will be the other, and ;r(io-;r) = 30 = lojr-r* ; 1 10 ± 1/ - 20 whence x=^ ^ ^ 2 = 5±l/-5: where the imaginary result |/ - 5 shows that there is some impossibility involved in the question. Upon examination we find that the largest product which it is possible to obtain from the two parts of 10 is 25. V. If 6 be zero, the value oi x reduces to •,jv In this case the roots are equal in value, but of opposite signs. The condition that this should take place is, then, that the coefficient of x in the first power shall be zero. If c be positive the roots are imaginary, but if negative they are real. Ex. 93. G'w&n {rs — aY-\-{rc—'bf — i =0 to find the conditions under which r will have values equal in magnitude but opposite in sign. Expanding, r\s^ + c^) — 2r{as + 6c) + rt' + i' - 1 = 0. .*. Condition is as + 6c = o, or, s c 49 b a The relations devploped in the present article are of the hijijhest importance in coordinate geometry, and in the appli- cation of algebra to geometry. 53. Limits 0/ positive and negative values of quadratic expressions. Let ax^ -t- 6;r + c be a general quadratic expression. Resolv- ing it into linear factors we find the expression to be ecjuiva- lent to 2a za Disregarding the factor a for the present, when the two factors within brackets have the same sign their product will be positive, but when they have different signs it will be nega- tive ; and the only effect of a change in sign of a is to reverse these results. But the bracketed factors can have different signs only when one is greater than ;?ero and the other less. Suppose the first factor to be greater than zero and the second one less ; then we must have ^> d + 1/^' - 4rtc 2a and <^:^-4^^ -£ 2a Between these limits for the value of x the expression is negative for positive values of a and positive for negative values ; and for all values of x beyond these limits the sign of the expression is the opposite to that which it has when the value of ;r is taken between the limits. Ex. 94. What are the limits of negative values for the expres- sion 3X^ + 2X- ^ ? Tu- • • 1 .. ^ / . 2 + V^4+ 60v , , 2 - V^4 + 6o V This is equivalent to, 3(;r+ ^ ) {x + 2 ), o o or 3(^+|)(^-i)- .'. X must be less than i and greater than -f, and if any quantity between these limits be substituted for x in the given expression the result will be negative. — 50 — Ex. 95. Under what conditions will 'jx - 3;r' - 2 be negative ? ' This is equivalent to -{ix^-']x^2)= -^{x-^^x-2). Hence the expression will give positive results for all values of X between 2 and \, and negative results for all other values* 54. Of maximum and minimum solutions of quadratic extressions. By dividing by the coefficient of x^ any quadratic may be put into the form, ^ x^-\-px-\-q-o. We know, Art. 49, that there are two quantities real or imaginary which when substituted for x in this expression will render it true. These are the roots. If, however, we put any other quantity whatever for x the expression will not be equal to zero, but to some finite quantity which we may denote by y. The value oiy will depend upon that of the quantity sub- stituted for x. If among all the quantities which can be sub- stituted for X there be one which will make y frreater than it can be made by substituting any other value for x, that value of ;r furnishes us the maximum solution, and y or the quadratic expression is said to attain its maximum. If on the other hand the particular substitution renders y less than any other substitution does, we have a minimum ^o\ut\or\ and v or the quadratic expression attains its minimum. 55. To find the maximum or minimum solution of a quadratic. Let x^-\-px-\-q=^y ; then,;r=rl±J^ZlS±4V. 2 Now, in order that x may be a real quantity the expression under the surd sign must not be negative. It is readily seen that increasing the value of y has no tendency to make p'i—^q^^y negative, and hence that y has no maximum. By diminishing jy howexer the value of the whole surd expression will be gradually diminished until it passes through zero and becomes negative. Hence y has a minimum limit ; that value which makes the surd expression zero. Again, let the expression be, -}^-\-px-Yq=y; i-atic, — 51 — t Changing signs, :i^-px~q— —y ; . whence ;r=^A!^2±SE47 2 In this case, since y is negative, increasing the numerical value oiy diminishes the expression under the surd ; hence y has a maximum limit when this expression becomes zero, but it has no minimum limit. We infer then, i. That every quadratic admits of a minimum or a maximum solution according as the coefficient of 3^ is respectively posi- tive or negative. ii. That the maximum or minimum solution is obtained by solving the equation for x and then equating to zero the quan- tity under the surd sign. In the general equation 3?-¥px-q-y^ the minimum value of jy is ii — ^-, and the corresponding value of ;r, or the value 4 of ;r which renders the expression a minimum is ^. Ex. 95. It is required to divide a number a into two parts such that their product may be a maximum. Let X be one part, and a-x the other. Then x(a - x) -y a maximum ; .'. ax~3?—y a ± /a' or x= 4y )ression seen make n. By )ression ro and t value Hence, for a maximum, a' - 4y = o, or y = t — \ , And the number must be divided into equal r.ud x = 2 parts. If a denotes a line, we see from this that for a given perimeter the square contains a greater are than any other rectangle. Ex. 96. To divide a given numberinto two parts such that the sum of their squares divided by their product may give a maximum or a minimum, and to determine it. Let a be the given quantity ; then x*+ia-xY = 2x^—2ax+a^ = sum of squares, and x{a —x) -ax-x^ = prod uct. 24^ — 2ax -\-'a^ .' ■ — = y = a max. or a min, ax -3? From this we obtain, 2 2 \ 2+y Whence we readily see that ^ can have a minimum value, but no maximum. Put I ^ =o .*. v = 2 = the min. value ; and x- — 2-Vy 2 Hence the number must be divided into two equal parts ; and the sum of the squares of the parts divided by their pro- duct cannot be less thah two. This article is of particular importance in the appli- cation of algebra to geometry. 56. Graphic representation of the quadratic. All the prominent properties of the quadratic may be ex- hibited graphically by means of a curve. Take for illustration the quadratic expression ;ir^-3;r- 2. We know that for two particular values of ;r, (the roots), the expression will be zero, but that it will have some finite value when any other quantity is substituted for x. Let y denote that value ; then x^ — ^x—2=y. Substitute different values for x, integers for convenience, and we obtain corresponding values of _y as follows : if ;r= -I o I 2 ^ 4. . . . . ^'^^ y= 2 -2 -4 -4 3 2 4 2 venience, -" 53 — Draw two lines xxj, yYi, inter- secting: at right angles in o. Let the different values substituted for X be denoted by distances meas- ured from o along xXj, the posi- tive values to the right of o and the negative to the left. Also, let the corresponding values oiy be measured from the linexxj parallel to the line yYi, the positive values upwards and the negative downwards. We thus get a series of points «, 6, c, d, e,f . . , . The curve which passes through these points and through all points similarly ob- tained by substituting all possible quantities for ;r represents the quadratic expression x^ — ^x—2. i. Consiier >he points P and Q where the curve cuts the lineXXj. t t icse points^' is zero, and they accordingly represent tht ...juation, x^- 2^+^ = o. And the values of x for these points i.e. OP and OQ, or the distances of these points from O represent the roots of the equation. We thus see that one root is positive and has a value between 3 and 4, and the other root is negative with a value between o and i. If both points, P and Q, were upon the same side of O the roots would have the same sign, positive if upon the right side and" neg.itive if upon the left. ii Since the curve actually cuts the line XX ^ the points P and Q are not imaginary but real, and the equation has con- sequently real roots. If the curve after approaching the line XX i turned and re- ceded from it without meeting it, the roots would be im- aginary. iii. Suppose that the curve merely touches the line XX^ at its lowest extremity M. This might be brought about by moving the curve bodily upwards : but in so doing the points P and Q would gradually approach one another and finally meet at the point of contact, and the distances OP and OQ would be one and the same. Hence this denotes equal roots. If the curve were still more elevated the points P and Q would become imaginary. ,\ — 54 — Hence we see that if a quadratic chan,t,'es its form con- tinuously so as to pass from real to ima:ijinary roots or vice versa, it must pass tiirouf^h the condition of equal roots. Compare Art. 52, iii. iv. As the curve lies' wholly below the line XX ^ from P to' Q, the quadratic expression X'-y--2 is ne^'ative for all values of x between these limits, and positive for all other values. V. Since the curve sweeps downward to a lowest point and then begins to ascend, the quadratic has a minimum value. If the curve were reversed and the apex turned upwards, it would denote the existence of a maximum value for the cor- responding quadratic. vi. If YY^ passed through M so that the cu.ve was sym- metrical with reference to the line YY-^, OP would be equal to OQ in magnitude, but would differ from it in sign. Hence the roots would be equal in magnitude, but opposite in sign. Art. 52, V. Ex. The quadratic 6 +;r- ;»r"-^ has equal roots, one positive and the other negative. It is positive for all values of a: be- tween the roots, and negative for ail values beyond them. It admits of a maximum but not of a minimum. The curve described as above is known in Geometry as the Parabola. Of the double solution furnished by the quadratic equation. 57. When the statement of a problem involves a quadratic equation, the two roots indicate in general two possible solu- tions to the problem ; the double solution being sometimes directly applicable and sometimes not. In purely arithmetical questions it usually happens that only one of the solutions is directly applicable, the other becoming so only after some changes in the wording of the problem. Ex. 97. A man died in a year A.D. which was ^^^ times his age : 13 years before the year was the square of his age. To find his age at death. Let ;tr=his age, then 33^^ .jr=:the year A.D. and 33¥^-i3=(-«^-i3)'' Herel the 3 Sill a neg^ Hei 55 IS the ition. Iratic aolu- fimes that pther the Is his To whence x=^6 or 3^. Here 56 is evidently the answer to the problem, but what does the 3^ mean ? Since 33iX3i - i3=if|i=(3^ - 13)" .•. ^^ satisfies the algebraical condition, but 3i - 13 = - 7|, a negative quantity. Hence we may interpret the two solutions as follows : \ since (after) the man was born 13' years ago the year A.D. 5 was the square of the years 1, c ^\ u • -^ betore the man was born .... 58. It sometimes happens in even arithmetical questions that both solutions are applicable. Ex. 98. A man buys a horse and sells him for $24, thus losinjj as much per cent as the horse cost in dollars : To find the cost. X \{x-. loss — X 24 the cost,-—- .X 100 Whence x=^o or 40. And •.• both soiiitioiis satisfy the condition, the problem is l(j a certain extent indeterminate. 59. In geometrical [)roblems and problems involving geo- ^ metrical mairniiudes, the double solution is frequently of the ihiKhest importance, and it should not be neglected, inasmuch as it often increases materially our knowledge of the problem in Ihaiid. [Ex. 99. The attraction of a {)lanet is directl/ proportional to its mass and inversely proportional to the square of its distance. The mass of the earth is 75 times that of the moon, and their distance apart is 240,000 miles. It is required to find a point in the line joining them where their attractions are equal. Let P be the point and denote ^ ^ pP by a;. Then PM = 240000 -at. e p m q [Attraction of 0=75 X 2 ! "^^ =lX and these are to be equal ; (240000 -;t)2 ' - 56 - hence 75(240000 -x)^=x^, whence, ;r=2i5i6o or 271330 miles. We thus see that there are two points of equal attraction, the latter of which lies beyond the moon at the point Q ; a result which, when once obtained, recommends itself to our judgment as true. 60. When a quadratic equation so involves a surd as to necessitate the process of squaring in the course of the solu- tion, it sometimes happens that the roots obtained are not those of the equation proposed, but of an equation differing in sign only from the original. Ex. 100. Given 3^+1^30 ;»: — 71 = 5 to finder. By the regular mode of solution we here obtain the values 4 and 2f for x, neither of which will satisfy the given equation, they being in fact roots of the equation, 3;r-v/3o;r-7i=5. In cases of this kind it is only by verification that wr' can determine whether we have a correct solution of the proposed equation or not. Again from the equation ^x-\- U^2x-2-j, we obtain x=i^ and x=i^, of which ;r=if only will satisfy the given equation, while ^=3 satisfies the equation ^x- \^2x- 2 = 7. The difficulty in these cases seems to arise from the fact that when we square a quantity we lose all trace of its oiiginal sign, and we have afterwards no means of determining alge- braically what sign it was at first affected by. Thus: V^2;r- 2 = 7-3;rand - l/2;r-2=7 - 3:1:, evidently be- come identical upon squaring, whereas they cannot possibly be satisfied by the same quantities ; so that any solution must give us either both roots belonging to only one of these equa- tions, or one root belonging to each. Whether any value of x can satisfy the equation ■^x + V3ox-yi=$ or not we do 'not know, but if there be such a value it cannot be found by the usual mode of solving a quadratic. is a — 57 — TWO OR MORE UNKNOWN QUANTITIES, 6i. It fre(}uently happens that the conditions of a problem require the introduction of mor ' than one unknown quantity in its statement. M>... * In such cases we require tor tuw complete determmation of the unknowns as many equations as there are unknown quanti- ties, and these equations must moreover be independent, that is, they must be such that any one of them cannot be obtained from the others by any legitimate process. The equations in such a set are termed simultaneous equations. Thus : x+2y-{-z = 8 is a set of three simultaneous equations involving the three unknown quantities x, y and z; and they are thus named because the values obtained for x^y and z must satisfy all the equations at the same time. This takes place when ;r=i,_y=2 and ^=3. 62^ If the number of independent equations be less than that of the unknown quantities, the equation can be satisfied by an infinite number of sets of values for the unknown quanti- ties, and the problem is said to be indeterminate. Thus if we have one equation with two unknowns, as 2;r-3^=io, it is evident that if we put any value whatever for x we can find a corresponding value for^. This species of equation is exten- sively employed in co-ordinnte geometry, where x denotes an abscissa of some locus Sindy ihe corresponding ordinate. 63. If the number of equations be greater than that of the unknown quantities, then some of the equations must be incom- patible with the others, or else they are dependent^ and hence redundant. Thus, if 3;r + a^' = 8 2x - y = 3 x + y = 1 be a set of three equations with the two unknowns x and y^ - 58 - the values which satisfy the first two cannot possibly satisfy the third, or those which satisfy the second and third cannot satisfy tne first, &c. ; i.e., one of the equations is incompatible with the other two. If the third equation were * + 3^ = 5, then since this may be derived from the other two, or any one of them from the remaining two, one equation is dependent, and, thus giving no new relation, is redundant. But if the equations are literal and are to be also compatible, some relation must exist among the literal co-efficients. Art. ^^. Ex. Ex. J LINEAR SIMULTANEOUS EQUATIONS— ELIMI- NATION. 64. When we so coml)ine two or .more equations as to get rid of a quantity we are said to eliminate that quantity between the equations; and the process of solving a set of simultaneous equations consists in eliminating the unknown quantities, one after another, until we finally have a single equation contain- ing only one of the unknowns. The methods of elimination will be considered under the following heads : 1. By comparison. 2. By substitution. 3. By cross-multiplication and addition and subtraction. 4. By indeterminate or arbitrary multipliers. 5. By determinant forms. These modes are all applicable in any case, but they are not all equa,lly convenient. Thus i and 2 are not often con- venient with more than two unknowns ; 3 may be applied to any number, and is one of the most practical ; 4 applies with greatest advantage to three« unknowns ; and 5 applies most profitably to three or more. 65. Elimination by comparison. This method consists in finding the value of the same unknown quantity or some func- tion of it in terms of the other, from each equation, and then equating these values. 66. tutef its vai Ex. I SI ■';ih satisfy cannot patible may be om the zing no patible, JMI- 1 to get between taneous ies, one jontain- der the tion. hey are ten con- plied to ies with es most nsists in me func- md then Ex. 99. Let be the equations. — 59 — sy . Then from the first, x=S--^'f and from the second, ;r=— +^. 5 5 . 8_?^^i8 42:. 3 5 5 Whence ^ = 3, and thence x=6. Ex. 100. Given — +- = a, ^+- = ^ tofind^-and^. X y X y mn n' , m' Here, — =«» ^mb- — . X y y nt^-n^ .•.-(w»-»>)=m6-«a,and>' = 4— -, y mb-na and from symmetry *= w'-n' ma—nb 66. Elimination by substitution. In this method we substi- tute for one of the unknown quantities in one of the equations its value drawn from another equation. Ex. loi. Given 4^±M=;r-y, and ?£z>' = l-2> to deter- 40 3 mine ;rand>'. From the first, ^-^^y = ^ox-^oy', whence, :« = 52.. And 4 substituting this for x in the second. 2. 5Z-J/ = ^-2y Whence ^=^, and hence x = \. 67. Elimination by cross-multiplication and addition and sub- traction. The following examples will illustrate this very im- portant method : 6o Ex.^102. ^ Let ax-by=o, x-\-y = cht the equations. Multiplying the second equation by b and adding to the first we eliminate y and obtain «M;+^;r—^; whence ;r= ^ «+*' and thence, y ^ ac a-^b Ex. 103. Let the equations be, 2* + 4J' + 5« = 49, 3* + 5^ -»- 6« = 64, 4* + 3^ + 4'«^ = 55 r . . . a . . .^ 2a - r 2t + r • • • • S^' -f 6-e = 43 d - 2y - 3

'-i5«-4M = 43 C ^ and d cannot hereafter be employed. Next eliminate u. C+2r Z2y-i$z-tx-$i, .... 2a- 7'....2i;r- 4y-4s = 77, . . . . 50+ « . . . . lye- z =72 To eliminate y ; g7 + 8(? . . . . 162:1; -47:5 = 669. .... A And finally, 47X-A — 543-* = 2715 ; whence«=5»J' = 4. * = 3. « = 2, /=i. 7 ^ — 6i — « Ex. 105. Given ax-^by = c, and a';r 4 ^'^^ -- c' to find_;rand>'. Multiply the first equation by d' and the second by d and subtract one product from the other, and we get, _ x{ab'-a'b) = b'c~bc. b'c — be X- ab' - ab ' ac — ac and from symmetry, j* =-7^ — ^. ab — ab We notice here that in order to eliminate y we multiply the first equation by the coefficient of y in the second, and the second equation by the coefficient oiy in the first ; and simi- larly to eliminate x ; hence the term cross-multiplicaiion. 68. Elimination by indeterminate cr arbiirary multipliers, f his method may be readily applied to the case of two equations, or to the case of three. Ex. 105'. Given 3x+y = y, and io;r- 2y = 2 to find x and^. Multiply one of the equations, the first for example, by the indeterminate multiplier / and add the product to the other, and we have, x{-^k + 10) +>'(A - 2) = 7^-1-2. Now this is necessarily true whatever value may be given to L But if k = 2, y will disappear from the equation and we obtain 164;= i6, or ;tr= i. Similarly if 3^ +10 = 0, ;r disappears from the equation and t here results, ^(-^-2) = 2-^; whence >' = 4. 69. If we have a set of three equations, for example: 2.^ — 3^ + -s^ = 2 + + 22f = I J' - ^x + 2j/ - 3^ = 5, it is possible to multiply them by such multipliers that when the products are added the coefficients of two letters may both become zero at the same time, and thus we may elimi- nate both letters at one operation. In the example given if we multiply the first equation by i, the second by -7, and the third by 5 and add, we obtain io;r=: 20. -- 62 — To investigate a rule for finding the proper multipliers. Let, ax H by -^ cz = d a'x + b'y + c'z - d,' a''x + b^'y + ^"^ = (V\ Then multiplying the first by /, the second by m, and the third by n and adding, we have, x{la + ma + ;/«'') +y{lb 4- ;;;6' • , nh") <■ ^(/f -f- ;«c;' + «c") Now if y and ^^ are both to disappear their coefficients in this equation must be zero. We must accordingly have lb-\-mb'^nb" = o Ic + uu' -| «c" = O Eliminating n between these, we ohtaiii / ;;/ ^,V' - b"c' b"c - be"' n and from symmetry each. = bc'-b'c' And having three equal fractions the numerators must be proportional to the denominators. Hence /, m, n may be any quantities proportional to b\ b"c\ b"c~bc'\ bc'-h'c respectively ; and these quantities themselves are usually taken as the multipliers. To apply this, notice, i. That the multipliers are made up solely from the coefficients of the letters to be eliminated. ii. That the multiplier for any line involves only coefficients belonging to the remaining lines. iii. That each multiplier is the difference of two products, these being formed of terms taken always in the same order. Ex. 106. Given, x- y-22^ >, 2x-\- y - T^z — ii 3x-2y+ 2= 4. We find for /, m and n, in order to eliminate y and z, the rs. and the n cien ts in ve must be to illy taken nade up ted. •efficients products, order. d z, the - 63 - I values -5, 5 and 5 respectively. Then multiplying and add- ing, we get 20;r- 60 and hence ^=3. Ex. 106'. Given, ax ^by - as~b{a-\-b) bx-ay+z = a-b x-\-2y - 2s = ^-a. To eliminate^ and e the multipliers are, l = 2a-2, m = 2b — 2a, n=a~b, whence we obtain, after reduction, x = a ', and similarly y = b and s = a-b. 70. Elimination hy determinant forms. If from the simultaneous equations, a^x + b^y ■\-c^z = d^ a^x^b.^y-¥c^z=d^, We eliminate J and z by Art. 68, or by any other means ^^ obtain for the value of x, ;r = '^^2f8_+'^2^3f 1 '^^3^\^2 ~^^1^3 ^'a -^2^1^'3 ~^8^2^1 ^i^a^a+^a^s^i +«3^i'"2 -^i^s^a ~^2^i^3 -«3^2^i The complex expressions forming the numerator and de- nominator of this fraction are determinants ; and as we ne they occur in the common process of elimination. The num- erator may evidently be obtained from the denominator by substituting d for a throughout ; and hence from the principle of symmetry in order to obtain equivalent expressions for y and z we must substitute d for b and c respectively in the above form. Taking the denominator then as the type form tic numera- tors may all be derived from it by substitution. In the case of three simultaneous equations involving three unknowns as above, each term in the denominator is the pro- duct of three «/^;;ie«/s or is of three dim .fusions. With four equations each term will be of four dimensions, and so on ; and determinants are thus divided into orders according to the dimensions of the terms. A determinant of the third order contains six terms, while one of the fourth order contains no less than twenty-four terms. For the purpose of denoting tliese expressions without writ- ing them in full the following notation is commonly employed : - 64 - denotes ^162 -^2^1 which is a determinant of the second order. Hi 61 Cj ^2 "2 ^2 *3 ^3 ^3 denotes aid^c^ +^^^2^3^! +''3^i^2 ~^i^3^a which is of the third order and is the same as «l(^2^3 -^3^2) — ^2('^1^3 -^3Cl) + ^3(^l^"2 -^2^1)' From this we see that «1 ^1 ^1 = «! *2 ^^2 -«2 <^1 Ci + «8 ^ Ci «2 ^2 ^2 <^3 ^3 ^3 <^3 ^2 ^2 *3 ^3 ^3 In Uke manner the determinant of the fourth order, rti bi Ci di = a^ «2 bi Ci di «3 63 Cs da at bi Ci di bi C2 di -«2 bz C3 dz bt^Ci di bi Ci di + «3 ba C3 ds bi ^4 di b^ Ci di bi Ci di bi Ci di a^. b^ Ci di bi Ci di bz C3 t/g These relations between determinants of different orders enable us to expand a given determinant, or to find its value. Ex. 107. To find the value of 342 I I 3 211 We have, 342 I I 3 211 I 3 I I 4 2 I I + 2 4 2 I 3 = 3(1 - 3) - (4 - 2) + 2(12 - 2) = 6 Ex. 108. 3123 = 3 021 -4' I 2 3 -h6 I 2 3 -7 123 4021 412 412 021 021 6412 301 301 3 I 412 7301 = 3(- 8+9) -4(1-8+3)4-6(2 -12) -7(3 -16) = 50 _i Cl C'i > K Cl d. ^2 Ci d. bi C4 d. -65- The follovvinfj principles established in works on deternii- nants assist us in the evaluation. i. If a column or row contains a common factor that factor may be placed outside and each element in the column or row divided by it. ii. Any column may be added to or subtracted from another column, or any row may be added lo or subtracted from another row without changinj^ the value of the determinant. iii. If two columns or two rows be exchanged the sign of the (letrrminant is changed. iv. If two columns or two rows be the same the determinant is zero. Applying these in evaluating the last determinant, we have, by bringing the third column first, which does not change the sign, it being a double exchange ; 3123 = 4021 6412 7301 2313 2401 1642 o 7 d, T- 2313 = 2 401 2 12 8 4 0731 I - I -2 8 i, 3 7 3 I ,1 4 2 o 5 4 2 -4 3 I = 2 I 3 I 8 1 - 1 8 7 3 2 I — 5 4 2 £ 2 5 -4 3 5 3 ■2 3 I by dividing the first col- umn by 2 and then sub- tracting the first row from the second and the second from the third ; by subtracting the second column from the first : by S'lbtracting the first row from the last and dividing by 4- = 2(5X5-oX3) = 50- Ex. 109. Given x = X -vy +e {b + c)x +{c + a)y +{a+b)2 bcx + cay + abe = o = o = I. 01 I c + a a + b ^ >_^ I ca ab III b + c c + a a+b be ca ab Now if in the second of these determinants we put ^ = c we — 66 — obtain two columns alike and the determinant becomes zero ; hence ^-c is a factor, and from symmetry a — b and c-a are factors. .*. the second determinant = - {a-b) {h-c) {c-a). But the first =6-c; ,. X = {a-b){c-a) {a-b)(a-c)' Similarly, y = z = (6-c) {jb-ay I {c-a){c-b)* o, o, 71. If we have a set of equations which do not contain a constant term, we can determine only the ratios of the un- known quantities to one another and not the unknowns them- selves. Let aiX tK b-iy + c^z - a^x + b^y + c^z = be a set of two such equations. Put — = w, -^ = «, and they become, z z. ayin + b^n -f Cj = o a^m -f 6a« 4-^2 = 0; and we see that the unknown quantities to be determined are m and «, i.e., the ratios oi x : z and ^ : z, or any other two ratios which we chose to fix upon. Now, m - — = - ■ ft, , C\ 61 K C\ H 62 = b^ C2 «1 b. «i bx ^2 b^ «2 ^2 X z bx b^ C2 b. Cx Cj, «1 by symmetry. Hence x, y^ z may be any quantities respectively propor- tional to the denominators. This result is practically identi- cal with that of Art. 68. 67- Ex. no. To find the ratios a . b \ c when x . y \ z ^ mb -\- nc - la \ nc •\- la - mb : la -\- mb - nc. X _ y _ e Denote the ratio mb+nc-la nc + la — mb la+mb — nc — by l_ V Then, -la + mb -^ nc — vx ■= o la - mb ■{■ nc — vy = o la -f- mb — r.c — vz = o and considering a, by c, v, as unknowns, we have a b c m n-x -m n-y m — n—z -I n-x I n-y l — n-z and by expanding the determinants, we obtain a b c I m- X l-m-y I m-z V -I m n I- -m n I m- -n V y 2mn{y+z) 2nl{z+x) 2lm{x-\-y) -/^Imn .'. a : b : c - mn{y-\-z) : nl{z-\-x) : lm{x-[-y). X 72. Of sets in which the number of equations is greater than that of the unknown quantities. In order that such equations may coexist there must neces- sarily be some relation among the coefficients. Thus if we are to have, ax + by = c bx -\- ay = 2c X + y — a -\- b + c, we must also have {a-\-b) {a-\-b + c) — zo ; and unless this relation exists the given equations cannot pos- sibly coexist. Let aiX + b^y-\-Ci = o, a^-^ b^y + Cj = o, flair -f ^aj/ + Ca = o be three equations involving the two unknowns x and j'. Eliminating y between the first and second, and then be- tween the first and third, we obtain, X - - cx b,\ ^2 *2 1 = _ Ci bi C3 bs «i bi 1 «8 bt (h bx > = mA;+/i, y=^miX-\-hi, jy = m2« + //2, are to exist together, determine the condition. Here, o I m h or, )n{hx - 112) + n?i(/j2 -h)-\- m^ih - //,) = o. INDETERMINATE ANALYSIS OF THE FIRST DEGREE. 73. As stated in Article 62, if the number of equations be less than that of the unknown quantities an indefinite number of sets of values may be found to satisfy the equations. Thus, \{ ax-\-by = c be the given equation involving the two quantities ;r and ^ we may evidently put any quantity what- ever for X and find a corresponding value for j/. In practice the number of solutions is restricted by the con- dition that the values of x and y must be positive whole numbers. Ex. 112. It is require dto pay three dollars in ii-cent pieces and 7-cent pieces. Let X denote the number of 1 i-cent pieces and y that of the 7-cent pieces. Then, I i^r -1-7^ = 300 is the equation. From this, ;r=3£^Lz7Z= 27 + ^ "^.T II II \ ;el eath ion of : UliX + hi, 1. IRST tions be number the two ty what- tbe con- /e whole it pieces lat of the -69- As ;r is to be a whole number, the expression ^ . ~'/ , and its II multiple by a whole number, must be a whole number. We now endeavor to multiply by such an integer that the coefficient oiy may be greater or less by unity than some multiple of II . 8 is such a number, since 8x7 = 56 = 5X11 + 1. Hence, -xIlAZ = 2 - 5^+ -^ must be a wh. no. II II and 2-y _ II a wh. no. =/> say. Then J/ = 2 - 11/), and putting this value in the original equa- tion we obtain, r=26 + 7/>. Hence, x=2b^ yp, y=2—i\p is the required solution, where p may be any integer, positive or negative, which will give positive values for x and y. If p = 0-1—2-3 ;f = 26 19 12 5 y = 2 13 25 37 which four setj- are all the possible positive integral solutions. Any other integral values for p would make either *" or _y nega- tive, which is not consistent with the original condition. Ex. 113. It is required to find a number which when divided by 3 leaves a remainder 2, divided by 5 leaves 3, and divided by 7 leaves 5. Let X be the number ; then, f , — 1^, ? must all be whole numbers. 3 5 7 Put ^^=/> .-. * = 3/' + 2 ; 3 and writing this for x in the second fraction, .3r ~ ^ must be a whole number. \ /. P ^ — q must be a whole number, .*. /> = 5^ + 2 and ^ = 159 + 8 ; i ' . — 70 — and this in the third gives, -M~^, or ? — ^= whole number = r ; 7 7 .'. ^y = 7r - 3 and ^=i05r- 37, where r may be any positive integer whatever. Making r=i gives 68 for the smallest number satisfying the required conditions. 74. If we have ax+dy=c an indeterminate equation of the first degree, it is readily seen that by increasing x, y may be made to pass through zero, and conversely by increasing y, X may be made to pass through zero. If then negative values of X and y are to be excluded, x cannot be greater than — nor a less than zero, and hence the number of solutions is necessari- ly limited. But i{ax-dy=c be the equation, an increase in the value of X must be accompanied by un increase in that of y, and as both may be indefinitely increased the number of solutions is quite unlimited. 75. In the equation ax±dy = c, a, b and c cannot have a common factor, for we may divide throughout by such factor and thus get rid of it. Again, a and h must be prime to each other, for if they have a common factor, it must also be a factor of ax±by ) but as it is not a factor cf c, the equation cix±by-c is impossible. Thus 2;r--|-ioj/ = 3i cannot have an integral solution, 76. In Ex. 112 we found for values of ;rand y, x=: 26 -\-yp, y = 2~iip. Now it will be noticed that the coefficient of/) in the value oix is the coefficient of^ in the original equation ; and sim'- larly the coefficient of/> in the value of _y is that of x in ti e original equation. This may be proved to be always the case.* Hence if ax + by = (' be the original equation, the values of x and y may be written, x = a± bp, y -i^ + ap, where a and (3 are fixed quantities, wliicli bolve tlic equation when p = 0. DemouBtrationB of tbiH kiud beloug to on aclvoiiced courBe of Algebra. Making required on of the ' may be leasing y, ;ive values an c a nor necessan- le value of y, and as :>lutions is )t have a ich factor they have but as it npossible. the value ind sirr"- in tie ways the ition, the ap, where ion when — 71 — If then one solution can be determined by any means, all the other solutions may be obtained at once. Thus, if we find one solution of Ex. 112 to be x- 12 and J/ =24, we have x-\2-\-jp, ^' = 24 -np as general formulae, and by making />= - 1, o, i, 2 successively we get all the pos- sible solutions. " If the equation be a;r-^^= c, we have only to change the sign of b in what proceeds. 77. In Ex. 113, the coefficient of r in the value of x is the L. CM. of the three denominators, 3, 5 and 7. Hence if / denote this quantity the value of x may be written, x=Y + lr ; and if one solution (y) can in any way be found, others will be obtained by adding on multiples of /. SIMULTA NEOUS QUA DRA TICS. 78. If an equation contains two unknowns, its degree is measured by the term of highest dimensions in these un- knowns. Thus, 2;»r+3;ry -1-4 = is a quadratic since the second term is of two dimensions. In like manner \fx,}f, z, be unknowns, ;jr2 -fj^ = o is a quadratic, ;r2j/ + 5^2 -(-^'^ = i? a cubic, xy'^z-\-z^ - 2xy = o is a quartic, &c. The most general quadratic in ;r and^ that can be written is, ax'^ -f bxy + cy"^ +dx + ey +f— ; and the most general in x, y and z, is cix"^ + dy^ -(- cz"^ + dxy -f exz +fyz + fx^- hy -j-kz + l =0. 79. In genera] the elimination of an unknown between two (juadratics produces an equation of a higher order ; but if one of the equations be linear the resulting equation will be still a quadratic. In any case elimination between two quadratics cannot pro- duce an equation of a degree higher than the fourth. As a consequence the solution of simultaneous quadratics may re- i is I — 72 — quire finally the solution of a quadratic only, or of a cubic, or of a quartic. Th« problem may, therefore, admit of two, three or four solutions depending upon conditions. Solution of simultaneous quadratics is often effected by in- genious combinations and artifices rather than by any fixed principles of elimination. These artifices are best learned by observation and practice. TWO EQUATIONS WHEREOF ONE IS A QUADRATIC AND THE OTHER A LINEAR. m li 80. The solution of these is effected by substituting in the quadratic the value of one of the unknowns as derived from the linear equation. Ex. 114. Given, att^ ^by^+cxy ■\-dx-\-ey+f=o, and }nx-\-'ny+p=o. From the second equation, x = - ^ — — . m And this value in the first gives, after reduction, yian^ + 6W - cmn) +y{2apn - cpm - dmn + em^) -\-ap^ — cipin +fm'^ = ; a quadratic in y. Ex.115. Given s^-2jf^-\-xy-y=i, and 2;r-3_y=i. Here, x = ^ — -, which in the first gives, 2 1(3^ + 1)^-2/+ -^ (3)' + i)->'=" 2 From which we obtain, v = i or — 4_ ; 45 and thence, ;r = 2 or - ^ .,' 25 81. If the quadratic equation be divisible by the linear the equations are equivalent to a pair of linears only, and x and ^ have but one value each. — 12^ — Ex. ii6. Given 3^''- 5^J'-2y= 17, 3 x-2y = i. The first equation is {x- zy) (3:r-f y) = 17. I^iit x-2y-\'^ .'. ^x+y=iy. Whence, ^=5,^ = 2. If we solve this l)y substitutinjj from the second equation in the first we obtain, x=i + 2}'; .'. 3+ 12^'+ 12^-5^- 12^=17 or ^ = 2, one value only. 82. Sometimes equations may be solved by combining them in some simple manner. Ex 117. Given x"^ + y'^ = 13 X + V = 5 Subtracting the first from the square of the second we have, zxy = 12; and subtracting this from the first, we jj:et {x -yf = I , or ;r -J/ = I ; .-. x = ^,y = 2.. SIMULTANEOUS EQUATIONS CONTAINING TWO QUADRATICS. 8-^. It is not always possible to solve tl.ese as quadratics, and experience is usually the only guide as to whether it is ()os- sible or not. Ex. 118. Given 2x'^ + ^xy = 26, 3/ + 2xy = 39. Here, 2X^ + ^xy = x {2X+$y) = 26, and 3/ -\- 2xy = y {2X+^y} - 39; V 2 2y .'. dividing L = -, and x = -=^. ^3 3 Putting this value for x in the first, — 74 — ^l + 2y'i = 26. 9 Whence, ^ = ± 3 and x = ± 2. 84. If the terms involving the unknowns he homopfeneous. we may advantageously obtain a third equation in which the unknown quantity is the ratio of one of the original unknowns to the other. Ex. 119. Given, x^ -^ xy -{■ 4^ = 6, Zx^ + 8/ = 14. Let — -V x = vy. Then ^'^+^+Ay'^ - '^'^y'^^'"y^+'\y'^ _ t;^-ft>+ 4 _ J_ 3^2 + 8' 7' whence we find, v = 4 or — ^ ; and writing x=^y in the second equation gives, }> = ±^, and .*. x= ±2. If we take the other value of v and write y= - 24: we obtain ;r= ±y^. ,, _ 2l/l0 ^ = + "^ 5 5 Hence x andy have each four values all of which satisfy the equations. 85. If the equations be each symmetrical with respect to the unknowns, it is frequently of advantage to employ two new unknowns, one of which is the sum and the other the difference of the original unknowns. Ex. 120. Given x^ +y'^ +x+y = 8, X +y +xy = ^. Put x=u + v, y = u-v ; then the equations become, 2_-,2 Adding, 2u^+s^ = 9 5 whence, w = f or - 3. — 75 — With thise values of u we find, { when «/ = f, v= ±^, x — z or i, y-\ or 2. when M=-3, i;=±j/-2, x- -^±\/ -2, y- -3+v^-a. Hence 4: and>' have each four values, which give four pairs satisfying the given equations. »n the present example, as in all cases where 4r and j' are symmetrically involved, their values are interchangable. 86. The substitution of the last article may sometimes be employed where the equations are not strictly s> mmetrical in X and^. u Kx. 121. Given, x^-^^x"^ -y^ = — Wx+y +\/ x-y ). (x+y)^ - {x-y^ = 26. Put x+y=2s^,x-y = 2t'^. The equations become, l/8(s8-/3)=26 ^ From a we get at once, or s3-/3-fs/(s_^)=8t/2 r Substituting for .s* - 1^ from /9 in y, we get s7(s - = -^ ^ 1/2 j9H-tf gives, ■ =— , St 3 • Art 27 (^ + ^)'-i6 (s -/)=» _ 4 . .*. —— = 2, and 5 = 31. s-t Whence we readily find, s = -^, « = — 7-, and V 2 V^2 hence x=^, and^ =4. ■' 1 1 76- INEQU.lLITIl'S. 87. An equation declares that there is equality hetween its two members, but a non-equation or inequality declares that one of its metnbers is greater or less than the other; and the problems which present themselves in inequalities usually require us to prove that one expression is greater or less than another. Since the square of a quantity is always positive, (x-y)'^ or x"^ ^y"^ —2.vy is a. positive quantity whether x be greater or less than y. Hence, x"^ +y^ is greater than 2xy ; or expressed symbol- lically. The proof of a large number of inequalities depends upon this principle. K x = j^ the inequality becomes an equality The following principles are important : If a > ^ nb, and a n > b J n but n a < n r 2. a + c > b + c, and a — c > b — c; but, c—a i/b and a" > b"^, but a' 2ab, b'^^c'^> zbc &c. een its es that md the usually :ss than -J/) 2 or ater or symbol- pon this luantity Id. — 77 — .'. 2rt*-|-2/'+ ... > 2ah-\-2bc-{- ... J .-. a^ hb^+ . . . >ah + bc-\ rd-i- . . . Ex. 123. I'or the same hasu and perimeter the area of an isosceles triangle is greater than that of a scalene one. Let s = ^ peri meTerot eacTiT^nd b= the commonbase. Also, let a, c be thf sides of the scalene triangle and e the side of the isosceles one. Then, ^, = area of isosceles = I s(s~e)\s -6), and A,^ " scalene - ]^s(s -a)(s-b)is -c). .'. A^^A^ AS {s—e)^^(s-a)(s-c). > < as e'-zsc -^ ac -s(a + c), as e' > < ac, since a-\-c = 2e ; ac. as, a' + c" ^ 2ac. But rt'-hc" > 2ac Ex. 124. x ^-\-y^ > x^-\-xv_^ . "I x^ ■\-y^ -^ x^y+xy^. as {x^-y^){x y) ^ o. But if ;r >^, both factors are positive and their product is positive and therefore > o. And if ;r < ^, both factors are negative and their product is positive and therefore > o. I::: ^ f: 1 ■ if s ■ 1 78 SERIES. 88. A succession of terms formed according to some regular law is called a series. If the number of terms be limited the series is finite, but if unlimited it is infinite. Series may be formed or developed in a number of different ways, one of which is given in Art. 9. Their study is important inasmuch as in many cases we are compelled to employ them. We have examples of what are the sums of the first few terms of well known series in logarithms, sines, &c. The law of formation of the terms of a series, or the "law of the series," may be very simple or very complex. The simplest series is one in which each term differs from the one before it by a constant quantity. Such a series is termed an equi-difference series, an arithmetic series, or an arith- metic progression. ^^ OF ARITHMETIC SERIES. -* 89. The quantities with which we have normally to deal in an arithmetic series are a, the first term ; », the number of terms ; d, the common difference between consecutive terms ; z, the last or n^ term ; and s the sum of » terms. Having any three of these we can find the remaining two by means of the relations which we proceed to develope. Let a, a-hd, a + zd, a + ^d, &c., be the consecutive terms of the series. Then it is readily seen that the n^ term is a-\-{n — i)d; .'. 2 = a+{n -i)d. . . . (A) To find S. S=a-\-(a-^d) + {a\qd)-\- , . . .(a + fT^.d). and reversing the order of the terms, S-{a + n — i.d) + ia + n — 2.d)+ .... +a • / adding, 2S= (2a +» - i.d) + {2a-i-n- i.d)+ .... to » terms, regular ted the may be , one of lasmuch Ve have of well irmation r be very srs from series is an arith- deal in imber of terms ; » two by e terms term is — 79 — = »(2«+«-i.i); .-. 5 =— (2« + «^ J) (B) 2 Formulae (A) and (B) involve all possible relations among the five quantities given above. Ex. 125. Given ^=13, the sum of n terms. Any three of these bein^ given th« remaining two may be found by the relations now Uj be devfeioped. t years r with a ■ the first t the end i+rt) :+r./-il • • • )receding \uimultiple ' are a terms, z of these 1 relations -83- Let a, ar, ar', at^, &c., be consecutive terms of the series. Then it is readily seen that the n*^ p^rm is (ir^'^> .'. z = ar^ ^. . . . (A) To find S. Multiply byr, rS= ar + ar'+ .... ar "•-j-ai'°"^ + «y* Subtract , S{i—r)=a—ar" J^^-Vv^ v-***^^^*** ar"" +ar' n-l «.n .'. ^~ci. . . . . {B) I ~r Otherwise as follows : By division, =rt+ar + «r'-}- I -r ar-'+-fn .-. s = a-\-ar + ay"-' = a ar" I - r I - r = a. r -r Formulae (A) and (B) involve all possible relations amongst the five quantities given above. Ex. 130. The population of a city increases at the rate of 5 per cent per annum, and it is now 20000. What was it 10 years ago ? In this case, since the series is a decreasing one r is a frac- tion, viz.: , a = 20000 and h = ii, as there are 11 terms to find z. From (A), z = ar'"' = ."^"""^"^ = 12422 nearly. 1.05 ,,., 20000 (1.05)^° Problems in Geometric series involving r or n as unknown quanties cannot in general be conveniently solved without logarithms. 96. If in (B) r is less than unity, r" may be made as small as we please by taking n sufficiently great. The liwii then to which s approaches as a becomes indefinitely increased is, , and this expression is usually taken as the sum of the in- I -r finite series in which r is less than one. It must be borne in *^ ^ ^ 84 - mind, however, that no number of terms which we could ever take would by summation be as great as , for as the num- ber of terms is infinite there must always be a remainder ; but by taking a sufficient number of terms we may make their sum approach the value of as near as we please while we can never make that sum surpass it. Ex. 131. To find the value of the repeater .36. This is equal to i^jfg- + nf^j^iy 4- ... ad infinitum. and s = a I —r = 36 _t. ^ T _ l \ ~ 3 6 y JJ) 0. - ol? n Ex. ^ J2. The series, i + — - + ft' For s = n-t I {n + I)'' »'■' n ^- 1 ■ + ad infin. =11 + 1 I - n n + i —n — n-i 1. n-\-i If « = i, 1+-^ + ^+ . . . . »=2, I+l + f + . . . . n=3, 1+I + A+ • • • &c., &c., &c. = 2, = 3. 97. In any three consecutive terms of a geometric series the middle term is called a geometric mean between the extreme terms. ., . Prob. To insert a geometricmean between two given terms. Let a and b be the given terms, and g the geometric mean required. Then, since a 7, b : re to form three terms of a geometric series, we must have -^ = — and a g g-=l/au Hence the geometric mean between two quantities is the square root of their product. (Compare Art. ^^ where it is called a mean proportional.) The side of a square is a geometric mean between the -85- sides of the equal rectanj^le. For if a, b be the sides of the rectangle, and s that of the square, area = fl6=s^ 98. Prob. To insert n terms between two given terms so as to fornn a geometric series. Let a and b be the terms, and let the completed series be, ^ n> ^a» 'Sj • • • • *n» *• Then, A-^'^'3_ But b_ a a tx U = ;'°+^ there being n-fi factors. ••■ -(7) ^» \"+i And 1^ n+l (h \"'r-i — ) = (fl"6)n I 1+1 U - ar^ ■■ \ a iy+' = (fl"-»62)n4.1 &c., &c., 99. If a sum of P dollars be put at interest for one year it amounts to P( I +>') dollars. If this be now taken as a new principal and be put at interest for another year it amounts to P{i +r){i +r) or Pii-i-r)'^. Similarly in three years it will amount to P{i+r)^ ; and in t years to P{i+ry dollars. Therefore if A denotes the amount we have A=P(i+jf. which is the fudamental formula in compound interest. It is evident that the amounts at the ends of successive years form the geometric series, P(i+r), Pii-\-r)^, P(i +r)3, . . . Pii+r)\ Ex. 133. n annual payments of P dollars each are made into a bank to remain at compound interest. To find the total amount due at the date of the last payment. Let R denote i +r. >-^^^. (T 86 The I St payment remains « - 1 yrs. 2nd n - 2 its amount is Pi?"**. Pi?"-». • • • (( last " " o " .-. /. The total amount is F{i+H-\- .... i^°-'), 7^" - I P. A=P. or K-i p (i + y)°-i This gives the amount of an annuity which has been fore- borne or left unpaid for a period of n years. To find the present value of such an annuity, or the sum which when put to interest will produce its equivalent, we have, ' 0- - /?» R^ (i+r)"' Ex. 134. A corporation borrows P dollars to be paid in n equal annual instalments, each instalment to include all interest due at the time of its payment. To find the value of the instalment. Let P denote the instalment and a, b,c, &c., the sums paid in successive years upon the principal. Then, ist payment =p = a + Pr, amount unpaid = P -a ; ^ 2nd payment =p = b + {P — a)r, whence b=aR, amount unpaid =P -a -b=P - a~aR ; 3rd payment =p = c-\-{P — a-aR)r .'. c = aR^, amount unpaid = P-a — b -c, &c. Similarly, n^^ payment =p = aR'"'^, * ' amount unpaid =P-a-b c- &c., = P-a-aR-aR^ - ... -aR'^'K But the amount unpaid after the last payment must be zero ; hence, P-a(i+i? + R2+ R"-*)=o, PR""-'. • • • P. sn fore- :he sum lent, we n n equal elude all find the ms paid aR, aR^, iR''-\ must be -87 - « - 1 fi^-i ii" - 1 Hence, 6 = -fl— + Pr = P.-!^. HARMONIC SERIES. 100. A number of terms is said to form a Harmonic series when the reciprocals of the terms form an Arithmetic series ; so that if the reciprocals of the terms be taken in any arith- metic series we have a Harmonic series. Thus I, 3, 5, 7, 9, is an Arithmetic serie?, and I, ^, ^, -f, ^, is a Harmonic series. Let a, b, c be three terms in Harmonic Progression ; I I I then a I T h I _ a . a-b^ a or a:c :: a are in A. P., and consequently, = the common difference. c b:b c. And three terms are in Harmonic progression or series, or they form a Harmonic proportion when the first is to the third as the difference 'between the first and second is to the difference be- tween the second and third. This is frequently taken as the definition of Harmonic Pro- portion ; and a series of terms in which any three taken con- secutively form a Harmonic Proportion is a Harmonic series. Problems in H. P. are best solved as problems of A. P. by means of the relation given in the first definition of a Har- monic series. Ex. 135. To find a Harmonic mean between A and B. Let H be the mean. Then, I -^, — are to be in Arithmetic proportion, ■m I H — 88 I I I ,2 I.I "'^aTh- J- /I V loi. Harmonic proportion is so nameil on account of the similarity which exists between its terms and the relative lengths of a trinj,' which sound the harmonics in music. Its chief application, however, is in Geometry. Let A, X, B, Y be four points \y i ^ in aline. Then AX, AB, AY form A— — X -^ — B Y three magnitudes which may be taken as terms of a harmonic proportion, if AX is to AY as the diffCi nee between AB and AX is to the difference between AYai.dAB; i.e., if AX: AY :: BX: YB. The points A, X, B, Y are then said to form a harmonic range, and the line AB is said to be harmonically divided in X and Y. The properties of harmonically divided lines is an im- portant one in modern geometry. VARIATION. If^'^%; I02. When 1 vvo quantities are so connected that a change of value in one is accompanied by a change of value in the other, in such a way t'lat their ratio remains constant, one of the quantities is said to vary as the other. Variation is usually denoted by the matk c/i , and is only a kind of geneialized proportion. A If ^ to B, then — ---constant = « suppose B .'. A =nB. Hence when one quantity varies as another they are con- nected by a constant factor. i. \i A sinB, A varies dir.fcctly as B. ii. Ifv4 = — , /I varies inversely as B. iii. If ^ =niBC, A varies jointly as B and C. C-c It of the B relative in music. B -Y AY as the e between harmonic videcl in X s is an im- a change lue in the nt, one of 1 is usually :eneialized -89- iv. {{ A— m --, A varies directly as B and inversely as C. . v./ Ex. 136. The space passed over by a body falling from rest varies as the square of the time, and experiment has shown that it descends 64 feet in 3 seconds. Find the relation between the space and the time.' S CO /'* we may write S = nt^. But when / = 2, 5 = 64. .*. 64 = 4» and « = 16. d its attraction AS the square iiuinher of beats Ex. 137. The earth's radius is 4,000 mil upon a body without it varies inv. of the distance from its centre. T which a pendulum makes in a day varies as the square root of the earth's attraction upon it. How much would a clock with a seconds pendulum lose daily if taken one mile high ? Let ^ = the earth's attraction at its surface, and r = the earth's radius. Then, g (o But if « = the number of beats per day at the earth's sur- face, and «i at the height of one mile, n c Co — .', n = — , where a is a constant; r r .". a=rn; and «i = — , where r^ =4001; are con- .'. n =n — ; and the loss = « - «i = «.-i — = 86400 x 4001 = 21.59 seconds. - -, A , . ^ nA 103. Let C vary as-77 ; then we may write C = -r^. B B Now if C is constant, A must vary as B ; and if B is con- stant A must vary as C. But multiplying by B, BC=nA; and therefore A varies as BC. pi ^. IMAGE EVALUATION TEST TARGET (MT-3) 1.0 1.1 IttlM 125 gKi 122 2.0 lU ■4.0 1^ ilM \m 1.6 '/a /. ^"j^ fliotographic Sciences Corporation 23 WIST MAIN STRIIT wnsTn,N.Y. usso (716)872-4503 K^ 5r X ^ % -~ I \ — go — Hence if ^4 varies as B when C is constant, and varies as C when B is constant, it varies as BC when both are allowed to change. Ex. 138. It is proved in Euc. vi. i, that the area of a triangle varies as its base when its altitude is unchanged ; and similarly it varies as the altitude when the base is unchanged ; hence it varies as the product of the base and altitude. If then A denote the area, 6 the base and p the altitude, we have {S ^ bp ; and hence A = nbp, where n is an un- known constant. Now the right-angled triangle whose sides are each i is one-half the square of which its hypothenuse is a diagonal, and therefore its area is ^ ; .•. « =^, and A =^^bp. If the three sides of a triangle vary so as to keep all their ratios constant, the triangle remains always similar to a given triangle. In this case /> c/» 6 and hence we may write p - mb, and therefore /^,^^mb^ ; i.e. the area of a triangle varies as the square of one of its sides when the triangle remains similar to a given triangle. PERMUTA TIONS— VA RIA TIONS- NATIONS. ■COMBI- 104. If a number of objects be taken and formed into groups such that the relative positions of all the objects are not the same in two groups ; then, |if each group contains al l the -Ob- jects concerned it is called a permutation^', but if it contains only a certain number of objectsTless than the whole, it is a variation. If the groups are such that no two groups contain the same assemBlage of objects, each group is calfea a combinations Frequently jip distinction is drawn between variations and permutations , and it is readily seen that the permutations are only the variations in a particular case. For this reason, and because the word variation has already been used in a different sense, we shall employ the word permutation for both. — gi — PERMUTATIONS. itains is a same [ready word 105. Take two letters a and b ; the permutations which ckn be made out of these are ab and ba, i.e. two. Take three letters and we have, abc, acb^ bac, bca, cab, cba, or six permutations. Similarly four letters will give us 24 permutations. But 2=1.2, 6 = 1.2.3, 24=1.2.3.4 ; From analogy we infer that with n letters the number of permutations is expressed by 1.2.3 ••..». io6. Let there be 4 letters a, b, c, d, and let us take only two at a time ; then we have, ab, ba, ac, ca, ad, da, be, cb, bd, db, cd, dc, or 12 in all. But 12 = 4.3. In like manner if three letters out of the four be taken at a time we would find the number of permutations expressed by 4.3.2. And if we employ 5 letters, taking three at a time, we have for the number of permutations, 5 . 4 . 3 or 60. Hence from analogy we infer that the number of permutations of n letters when r are taken together is expressed by «(« - i)(«- 2) .... to r factors. We propose to show that both of these inferences are correct. 107. Let a, b, c . . . . n he n different letters, and let us adopt the symbol nPm to stand for " the number of permuta- tions of n letters with m letters in a group." (i.) If we place only one letter in a group we can evidently have n groups and no more ; .*. «Pi =«. (2.) Put a aside and we have « — i letters left; and these taken in groups of one give « — i groups. Now place a before each one of these letters, and we have m-i groups of two letters in which a comes first. Similarly by operating on 6 we will have n - i groups of two in which b comes first ; then «- 1 in which c comes first ; and so on. But there are n dif- ferent letters to come first, and each of these gives us « - 1 groups ; .'. the whole number of groups of two letters will be «(«-i). .'. nF2-n{n-i). \ : '1^^* 92 — (3). Setting a aside again we have « — i letters left; out of these taking two at a time we may form (« — i) (« — 2) groups. For if n2r = n{n—i), then («- i)P2=(n - i) (« - 2). Now put « before each groug, and we have (n-i) (» — 2) groups of three letters, with a first ; and a like number with b first, and with c first, &c., and as there are n letters to stand first the whole number of groups is n{n - i) (n - 2). .*. «P3=«(m- i) («-2); and the law is manifest. Suppose this law holds for r things in a group, then nPr = «(» - i) (« - 2) . . . . to r factors. Putting a aside we have » - 1 letters, and these taken r together give (« - i)P/' = (« - 1) (» -2) . . . . to r factors. Now putting a before each group we introduce an additional letter and thus have r-\-i letters in a group. Hence there are |(«-i) («— 2) .... to y fact.[ groups of y-f-i letters with a standing first. Similarly there is the same number with b standing first ; with c standing first; and so on. Hence there are«|(«-i) («-2) . . . .' r fact.[ groups of r-f-i letters alto- gether. Or nF{r-\-i) =n{n - i){n - 2) . . . .r + i factors, since we introduce the additional factor n. If then the law holds for n letters taken r together, it holds when taken r + i together. But it holds when r = 3, and therefore for r-\-i or 4, also for 4+1 or 5 and so on for any number. .*. generally, nPr = nin-i) jn-z) . . . . (n-r-{-i) . -r^ Making r equal to n we have for the number of permuta- tions of » things when taken all in each group, tiFn, or simply P =w(» - i) (- ;) = 1.2.3 ••••»• 3-2.1 108. The continuous product of « consecutive natural num- bers beginning with i is called factorial n, and is indicated by the symbol «! or | n. Thus 4! or | 4 means i .2.3.4. Taking the formula nFr = n{n- i)(»- 2) . . . . (»-r + i), and multiplying and dividing by i . 2 . 3 . . . . (» -r - i)(» -r), we have, t; out of ) groups. I) (» - 2) er with b to stand then Mpr e taken r rs. Now inal letter there are rs with a sr with b ;nce there iters alto- ■, it holds = 3, and for any permuta- iral num- icated by 3 •4- + i), and -r), we 'jv -vn /l-z O"*""^ '^tJt^' -HUx-^M^'^ 93 „Py- »*(«-i)(»-2) {n-r^i){n-r) ....3.2. 1 .! I . 2 . 3 . . . . (» — r) n or «Py = — = . . . . B. n-r Making r = « in Art. 107, A, we have for the number of per- mutations when all the articles are included in each group, P = n. n But making r = « in B, we have P = , — . o Hence we must interpret | o as meaning unity. Ex.139. If »P4:(« + 2)P5::3:56, find n. «(» — 1)(«-2)(«-3) — TS 5 (n 4- 2)(» + i)n{n - i)(m - 2) ••• 56(w-3) = 3(« + 2)(« + i) ; whence « =6 or gf. Of which, although both numbers satisfy the condition, the integer only will apply to articles. log. // u of the article^i be alike. If the u articles were all different they would give rise to | u permutations, each of which could be combir.ad with each permutation from the re- maining articles, and this would give the complete number of permutations of n different objects taken all together. If we denote the number of permutations of » articles taken all together, of which u are alike, by P(w) we have P(w^ . |_w=P= |_«; and .'. P(m) = Similarly if v other articles be alike, P(«) (v) = _ '— . n u r u V Ex. 140. How many permutations can be made from the letters in the word Ontario ? Here « = 7 and « = 2, since there are two O's ; |.-. P(2)= 7:5:5^^^ = 2520. // |fi-.i ill :1 M 11 i III 94 COMBINATIONS. .1 ^' 1 10. Let «Cr denote the number of combinations of n things taken r together. Then from the definition of a combination each one would give rise to | r permutations. For abed forms only one combination however you arrange the letters, while it can give 1.2.3.4 different permutations. lIHence, (tfcLe^nunL- ber of combinations} X (the number of jgermutations whi ch can ^ made from eac^h^ combinat ion ) = tITe totaF number tations that is, nCrX I r = n n -r ; Art 108, B. n '. nCr = ii- \ r \ n-r r • • • • v^ This may be put in another form ; I n _ n{n - i) . . . . («->'+i)(» ->•) .... 2.1 n—r (» -y) .... 2.1 = n{n- i) . . .. . (n—r + i) = n{n— 1) .... to r factors ; And I r = 1.2.3 • • • • to r factors ; . to r factors . . . . D. r- n n-1 n — 2 .if From this it appears that the product of any n consecutive integers is divisible by factorial m, since »Cr must necessarily be an integer. Ex. 141. How many different guards of 4 men can be chosen from a company of 1*0 men ? Here « =? 10, r=4 ; .*. ioC4 = — .^ . - . ^ = 210. 12 3 4 III, If in Art. no, C, we make n — r—p, we have r=n-p, (^■..V 9: llo- (V^vvyt^ -'Jll_ ^— ''^/--'/j K^.S. Yi I n things nhinaiion *cd forms rs, while, lie_num: ;hi ch can if permn- .D. nsecutive ecessarily V 95 and nCn —p = n n-p\P and substituting r for p, ' n nCn -r = I r I n — 1 nCr. )e chosen j . J r = n-p, ^JL-Yl 1 V Hence the number of combinations of n things taken r to- gether is the same as that of « things taken n—r together. This must necessarily be true for the following reasons : — When from n things we take out r to form a combination, we leave another combination of » - r things, and therefore the number of each must be the same. These are called supple- mentary combinations. Thus 6C2 = A.A=i5 : 664= — .-5-.-4..J_ = i5. 12 1234 112. Forming the combinations of 6 articles i at a time, 2 at a time, &c., we have, 6Ci = 6, 6C2 = i5, 603 = 20, 604 = 15, 665 = 6. Hence if n is an even number the largest number of combinations ft can be made by taking — articles at a time. Again, forming the combinations of 7 articles i at a time, 2 at a time, &c., we have, 7Ci = 7, 702=21, 703 = 35, 7^4 = 35» 7^5 = 21, &c. Hence, if n is an odd number the maximum number of combi- nations occurs when the articles are taken or — I^ at a time. 2 2 In this case there are two greatest terms. 113. To find how often any one thing occurs in the combi- nations of n things taken r together. If from all the combinations containing a we take out a we will have left the combinations of » - i things taken r—i to- gether. Hence in the combinations, «0r, any one thing occurs n - iCr—i times. Similarly any two articles will occur together n - zOr — 2 times, &c. -96- li: ■ H i i'! 'f Ex. 142. The number of combinations of n letters 5 together in all of which a, b, c occur is 21. Find the number when taken 6 together and in all of which a, h, c, d occur. ^ 117. Dividing both sides in C, Art. 116, by | n, we get, ![i±fl°=^-i- n n |« — I ' I r l» - 2 — -fp^ — -i- D. I 2 «-3 3 A fourth and very symmetrical form of the theorem. 118. We have drawn these expressions for the Binomial theorem from the expansions of (i 4-;r)* and (i +;•;)'. We shall now prove that if the theorem is true for (i +4:)", it is also true for (n-;r)'^+^ Putn-i-i=w, then n = m-i; and writing this for n in B, Art. 115, 1.2 + (w-i)(w-2)(m -3)y3_|_ I . 2u 3. Multiplying both sides by i ■\-x, using detached co-efficients ; 1.2 (w-i)(w-2)(m-3) + ^ — 1.2.3 + ; x(i + i) 1.2 1.2.3 Hence the formula is true for m ; and m = n-\-i, .'. &c. But the formula is true for 71=4 as we have seen, .•. it is true for » = 5,' and if for « = 5 then for m = 6 and so on ; i.e., it is generally true when n is any positive integer. We have thus proved that the Binomial theorem holds when n is any positive integer. It may also be proved that it holds when n is any quantity whatever, but the general proof is be- yond the scope of this work. 119. The following generalizations are readily drawn from the form of the theorem. i. If « be a positive integer the series is finite and consists of » + 1 terms, •.* n terms contain x and one term is without x. 01 em IS . . . D. Binomial We shall also true n in B, tfficients ; &c. .•. it is ; i.e., it ds when it holds of is be- .wn from consists ithout X. — 99 — ii. If n be not a positive integer the series can never termi- nate, as reducing n by units can never give a factor equal to zero. iii. If n be a positive fraction and x negative, all the terms after the first are negative. . iv. If n be negative and x negative, all the terms are posi- tive. Ex.146. /T+^=(i+:r)^=i+i^+i^^^*'-f- ^ X ^' 2"^ ~~' ' " ' •• • ' ■■ — ■ • ^ i~ • • • • 4 X.2 = , + ^-i .(?)' + -L3.. (?)•- + .... 2 1.2 ^2^ 1.2.3 ^2' Ex. 147. ^ ^ ^ 2 I.2V2'' I.2.3V2' Ex. 148. I -X 1.2 1.2.3 = i+x+x^+x^+ .... (Art. 9.) Ex. 149. a^/ ( \a/ 1.2 ^a^ = a~ 2X^ • • • 3a* I.2.3^rt'* 120. The Binomial theorem may be used for the expansion of the power of a trinomial or polynomial. Ex. 150. {i-^ax-\-bx^)^ = i-^n(ax+dx^) + ^^^^^^^{ax + bx^)^'\- 1.2 = I + nax + nb n(', nin-i )^^ 1.2 ;r2'+«(»-i) 2ab + 1.2 n(n- 1)0; -2 )^8 1.2.3. a* ^ "y" • • • • !l IOC) Ex. 151. }^l+X+X^:s{l+X-\-x'*y =» I -t iix+x^) -f ilri^ [X x"^ )'* 1.2 ^J^~8 re [X x')' f . 121. The Binomial theorem may sometimes be employed to approximate to the roots of numbers. Ex. 152. Required the fifth root of 12. 12 = 32 -20 = 2*(l -ft) .-. l/l2 = 2{l-ft)^ = 2|l-^.|- -t^(ft)2- ±1^.{^)^-. . . .} 2-5' 2.3-5' = 2|i-i -^-jh- } =i-^5 nearly. LOGARITHMS. We propose to deal here with the nature and use of Logar- ithms, and not with their development. 122. Take the equalities, 2° = I, 2^ = 2, 2^ = 4, 2^ = 8, 2*= 16, 2" = 32, 2' = 64, 2^ = 128, &c. ; the quantities i, 2, 4, 8, 16, 32, 64, 128, &c., are numbers ; the indices of 2, i.e., o, i, 2, 3, 4, 5, 6, 7 are the corresponding; logarithms, and 2, the number raised to the several powers, is the base. By tabulating these, as in the margin, we have a table of logarithms to the base 2. In like manner we may form a table of logarithms to the base 3, or to any other base which one may choose. For common purposes the base employed is 10, for being at the same time. the base of our numeral system, it possesses certain practical advantages over every other number. To illustrate the practical applications of logarithms we may employ a table to any base TABLE. No. Log. I 2 I 4 2 8 3 16 4 32 5 64 6 128 7 &c. &\i. lOI — wliatever, for Ihc general properties of lu<^arithins are the saute fbr all bases, 'lakinjj tlie table above, then, let it be required (i) to multiply i6 by 8. log. 16 = 41 . J m ployed Number of which 7 is the log. = 128, .*. 8X16= 128. (2) To divide 64 by 4. log. 64 = 6 log. 4 = 61 btract • • • • r if Logar- 2«=l6, 16, 32, > 3» 4» 5» )er raised fVBLK. Log. O I 2 3 4 5 6 7 Ac Number of which 4 is the log. = 16, .•. 64-^-4 = 16. We thus see that imiltiplicHtioii of numbers corresponds to ailiiitjon of logarithms, and division of numbers to subtraction of logarithiis. This will be shown more geneially hereafter. 123. The above table is not complete, even as far as it goes, sirce the numbers do not follow each other in order. Thus it lav'ks the numbers 3, 5, 7, 9, &c. To find the logarithm of ono of these numbers we notice that the numbers in our table are in geometric progression while the h garithms are in arith- metic progression. Hence the geometric mean between two numbers must correspond to the arithmetic men between their respective logarithms. Thus 3] is tli» logarithm of I 8X16 or II 3136 .... This may be readily shown as follows : 27=128 = 8x16; .-. 2''^'"- 1/8X16, (jr 2=*^» = 11.31 .. . .'. 3i = log. II. 31 . . . By this means we may calculate the logarithm of 3. 1. I 2X4 = 2.8284; i(i + 2) = i.5 .-. 1.5 = log. 2.8284, 2. J 4X2.8287=5-6568 : ^(2 + 1.5)= i.7j = l-'g- 5-6568, 3. 1 '2X5^6568'= 3.363 ; ^(1 + 1.75) = 1-375 = lpg- 3.3630, 4. 1/2.8284 X 3.3630 = 3.0842 ; ^11.5 +1-375) =1-4375 = log. 3.0842 ; And by continually approximating towards 3 we at last find 102 — « ■ log. 3 = 1.585 . . . approximately. And in this way, although exceedingly operose, the logarithms of the prime numbers were once calculated. We infer then that 2^ **^=3, i.e. 2**^"" =3, or 2 1586 _ olOOO Of course we have no means of proving this except through logarithms themselves. 124. The Base. In the computation of logarithms by means of series, we come naturally upon a system having the strange number 2.7182818 . . . , j^enerally designated by e or e, as a base. These are called natural logarithms, Napierian logar- ithms, and sometimes hyperbolic logarithms. This system is usually employed in mathematical analysis. The only other system in use is the one having: 10 as a base. These are common or decimal logarithms. Let a denote any base ; then, *.• a" = I, the logarithm of i is always zero. If rt > I, then ci" > i^and «"" < 1. And, since a is greater than r in both systems of logarithms, the los^arithm of a qiiantiiy greater than i is positive, and of a quantity less than i, nef^a'.ive. Thus log 3 is a positive quantity ; but log .3 is a negative " Since a'* = — ^ = o .*. log. o = -00. Hence the logarithms of all proper fractions lie between o and - 00 . And since fl* = 00 , the logarithms of numbers above unity lie between o and + 00 . Since if a is positive no power of a can be negative it fol- lows that negative quantities have no special logarithms. 125. The number which we found for log. 3 to the base 2 is composed of two parts, an integer i called the characteristic, and a fractional part .585 . . . called the mantissa. In decimal logarithms the distinction between these parts is important. 126. Ihe characteristic. Since, io"^ = .ooi, lo"^ = .oi, 10'* s= .1, 10® = I, 10^ = 10, lo^ = 100, 10^ = 1000, &c., we have, number, .001 .01 .1 i 10 100 1000 &c. logarithm, -3-2-101 2 3 &c. although bers were lis except by means le strange r €, as a ian logar- i analysis. LS a base. ogarithms, , and of a logarithms nd since between o ve it fol- base 2 is iracteristic , lese parts .01, 10 have, 1-1 [C. EC. — 103 — We see from this that the characteristics are the logarithms of numbers made up of unity and ciphers only. Also, for a number between 100 and 1000, log = 2 + a decimal 10 •' 100, log = 14- I " 10, log = + .1 " I, log= -1 + " .01 " .1, log = -2+ " &c. &c., &c. Hence we may write down the characteristic of the logar- ithm of any given number at sight by the following rule : // the number is a decimal the characteristic is negative and greater by unity than the number of ciphers to the right of the deci- mal point. If the number is integral or contains an integral part the charac- teristic is positive and less by unity than the number of figures in the integral part. Or by the following rule : Call the units place zero and count from it to the significant figure farthest upon the left. The number of that figure is the character- istict, positive if counted leftward, negative if rightward. E.x. 153- To find the characteristics of, .00000734, 386.5, 943007.0162. 123456 210 643210 0.00000754 386.5 943007.0162 units place. .-. -6 units pi. units pi. ••• 5 For reasons now readily seen the characteristic is not usual- ly written in tables of common logarithms. 127. The Mantissa. Let log 425 be 2+m, where m is the mantissa or decimal part. Dividing 425 by 10, we must subtract the log of 10 from that of 425, (Art. 122). .•. log of 42.5 = 1 + w. Dividing by 10 again, log of 4.25=0 + w. Dividing by 10 again, log of .425 = — 1+ in. &c., &c, it — 104 — We notice that the mantissa remains constant, the mi'y change being in the characteristic. Hence we may sum up the significance of the parts of a lopfarithm as follows : The mantissa is connected with the group of figures and their arrangement; //i^ characteristic, with the position of the decimal point. 128. A table of decimal los:arithins re;^nsters only mantissae ; and since these start from zero at every power of 10, the table extends only bi^tween two consecutive powers of 10. For 7-place logarithms, i.e., for those with 7 decimals in the man- tissae, the usual extent is from 10* to 10'. We give below a portion of a table of 7-place logarithms taken from Hutton's tables as published li\ Chambers. No. 1 2 3 4 5 6 7 8 9 D. 2397 379C680 CG86 7043 7224 7405 7586 7707 7918 8130 8311 181 98 8492 8673 8854 9oa5 9216 9397 9578 9759 9940 0121 99 3800302 0484 0665 0846 1027 1208 1389 1370 1750 1931 2400 2112 2293 2474 2655 28:^0 »)17 3198 3379 3500 3741 01 3922 4102 4283 4464 4645 4826 50O7 5188 5368 6549 02 5730 5911 6092 6272 6153 6631 6815 0095 7176 7357 03 7538 7718 7899 8080 8261 ai4i 8622 8803 8983 9104 04 9345 9525 9706 9887 0007 0248 0428 (kiOi) 0790 0970 OS 3811151 1331 1512 1693 1873 2031 2231 2415 2595 2776 D H 181 P. 18 36 54 72 91 109 127 145 163 129. The workings; of a table of loj^arithms consists in two opera- tions the converse of one another, viz : (a) ^iveii an arrange- ment of figures to find the cories[)ondinLr mantissa, and (6) given a mantissa to find the corresponding arrangement of figures ; for the characteristic n'>t being registered has no im- mediate connection with the table. (a) Given an arrangement of iigures to lind the correspond- ing mantissa. The table above mentioned gives the mantissae of all arrange- ments of 5 figures at sight; lour of these are found in the column of numbers marked No'., and the filth in the horizontal row at the top. i. When the arrangement contains 5 figures. To find the mantissa of 23987, start at 2398 in the first lie at) y sum up nd their decimal ntissae ; he table o. Fol- ic man- arithms 9 D. Ul 181 lai )31 741 349 157 104 )70 m LC3 o opera - ;lnall?e- and (6) ment ot no im- lespond 111 range- Id in the Irizontal he first — 105 — column and pioc^e orizontally until in the column marked 7 at the top. To t. . figures 9759 there found prefix the 379 which the first colunin shows to be common to several rows. We thus have, mantissa of 23987 =3799759. ii. When the arranf;ement contains leas than 5 figures. Add ciphers or suppose them to be added to raise the num- ber of figures to 5, and then proceed as in i. Thus, mantissa of 24= mantissa of 24000 =3802112. And, log 24=1.3802112. iii. When the arrangement contains more than ^figures. To find the mantissa of 2403872. We find the mantissa of the first 5 figures to be 3808983. In order to show what is to be done with the remaining figures 72 we shall explain the column and row of the table marked D and P respectively. Mant. of 24038 = 3808983) D = difference of " " 24039 = 3809164) mantissae= 18 1. Now the 72 occurring here is -^-^ of the difference between 24038 and 24039. .'. we should add to 3808983, 3^X181. But33jrVXi8i=7XW + iV(2XW). The row marked P (proportional parts) gives the multiples of ^j^ from I to 9. Thus under 7 at the top we find 127 which is 7X-^]^ to the nearest unit. Under 2 v.e have 36, one-tenth of which is 3.6 or 4 to the nearest unit. Hence the mantissa of 2403872 is 3808983+127+4 = 3809114. iv. As a special case let it be required to find the mantissa to the arrangement 24044. Referring to our table we find the first cipher overlined 0~. This indicates that the three figures to be prefixed to the four there given change at this point from 380 to38i. The mantissa is accordingly 3810067. (6) Given a mantissa to find the corresponding arrangement of figures. Take the mantissa 3806745, for example. The highest mantissa in the table capable of being sub- tracted from this is 3806634 ; and we proceed as follows : Ill '■ *■ \ — io6 — Mant. given . . . 3806745 tab. mant 3806634 arrrang't = 24025 Diff. of mant. highest subtractive number from P . . . . Ill, 109 number 2, I. Diff. . . . Subtract number from P after dividing by 10 . . . 1.8 .. . number . .'. arrangement = 2402561. 130! To find the logarithm of a number. Find the mantissa of the arrangement without any reference to the decimal point, and then prefix the characteristic according to rule Art. 126. To find the number answering to a given logarithm. Find the arrangement corresponding to its mantissa and then fix the decimal point by means of the characteristic and rule Art. 126. t 131. To perform multiplication by logarithms. Let a" = m, then x = log m ; «*'=«, then _y = log «. Then, mM = a*.a^ =«*+^; and log mn=x+y = \ogm-\-\ogn. .'. to multiply numbers we add their logarithms and take the number answering to their sum. Ex. 154. To multiply 23.974 by .024056. log 23.974 = 1.3797405 log .024056 = 2.3812234 log .57671 = 1.7609639, sum. •'• -57671 is the product. As in this example, the negative sign of the characteristic is placed above it to save room, and it must be borne in mind that although the characteristic may be negative the mantissa is always positive. 132. To perform division by logarithms. with the notation of Art. 131, 107 — nantissa lal point, t. 126. Find the n fix the ule Art. take the eristic IS in mind mantissa m n av tn and log — = x -y sslog tn -log n. n .'. To divide one number by another subtract the logarithm of the divisor from that of the dividend and take the number answering to the difference. Ex. 155. To divide 1.4936 by .007453. log 1.4936 = 0.1742343 log .007453 = 3.8723311 log 200.4025 = 2.3019032, difference. .'. 200.4025 is the quotient. 133. To raise a number to any power. a*=» ; .-. nv =(,a'^)y=a'y \ and log nv = xy =_y log n. .'. To raise a number to any power multiply the logarithm of the number by the index of the power required, and take the number answering to the product. Ex. 156. To find the 21'* power of 1.06. log 1.06 = 0.0253059 21^= index. log 3-39957 = 0.5314239 .-. (1.06)21 = 3.39957. Ex. 157. Find the value of (.4726)*. log .4726 = 1.6744937 8 log .00248857 = 3-3959496 .-. (.4726)* = .00248857 . . . In this example the mantissa being positive we have, upon multiplying, -8 + 5.3959496= -3 + -395 • - • 134. To extract any root of a number. Since a" =«, 1 1 « («)!' = (a*) v = «v. \ - io8 '■ y .'. To extract any root of a number, divide the logarithm of the number by the number denoting the root to be extracted, and take the number corresponding to the quotient. Ex. 155. Find the value of (.017325)4^ log .017325 = 2.2386732, Divide by 7 gives, i. 7483819, corresponding number = .56025. In this case having a negative characteristic we make it evenly divisible thus : 2.2386732=7 + 5.2386732, which divided by 7 gives the quo- tient found. This is the equivalent to — Hf + .034 . . . = - I +^+.034 . . . = - I +.748 ... Ex. 159. To find the value of —5^ '- — 7;, being given the logs 2^X(2l6)* of 2, 3, 5 and 7. 1 1,3 Numerator = 3^X3^X7^X.o5<'=3 2 X7^X.o5« ; .-. log num. = Ji^ log 3+6 log 7 + 6 log .05=0.36569. Denominator = 2^ X (2^ X3^r =2^. X2^X3^ = 2X3*; .*. log denomr. = log 2 +f log 3=0.58730. .*. log of the value = .36569 - .58730 = 1.77839 ; and value = .60033. Those who make very great use of logarithms, as astrono- mers and navigators, do not usually employ negative indices for the logarithms of fractions, but make use of a system much more convenient in practice, although probably more difficult to master at first. An explanation of that system, as well as of other conven- tions in logarithmic practice, can scarcely find a place in this work. ■t:-;i u — log — im of the and take make it s the quo- ;n the logs 6 . > ,569. = 2X3^; astrono- [e indices a system jbly more conven- [ce in this EXPONENTIAL EQUATIONS. 135. An equation in which the unknown quantity is involved as an index or exponent is called an exponential equation. These usually require the application of logarithms in their solution. Ex. 160. In how many years will a sum of money double itself at 3 per cent, compound interest. From Art. 59, A =P(i+r)*. But A=2P .'. (i+r)* = 2. And going to logarithms, t log (1.03) = log 2, .'. t = - log 2 log I. 03 =^3+ years. Ex. 161. Given a'^ + a'^ = b to find x. Multiplying by a', a"^' - ba^= - i ; . , b±y¥^ .'. a'- And X'- log((!) ± v^6'- 4rt) - log 2 lot? a n^ X *W5— a ml =6 to find X a ) nx — lo^ a = {nx — a) log b ; m •. x(—\o^ a-n log b\= —a log 6 ; X = ma log b »(mlog6-loga) ° 1 — no CONTINUED FRACTIONS. iq6. Let us take for illustration the fraction^^, "^ lOI Then, 45 _ i _ i lOI lOI ^ , II " 24- 45 45 Again, ii _ i _ i 45"5 4 + ^- II II .*. 45 _ I lOI 2 + This latter expression is usually written i or more 2^- ± ^ II compactly- - — and ia called a conliniied fraction, which is ^ •^2 + 4 + 11 rational when the number of terms is limited, and irrational when not limited. 137. Toconvert any fraction into a continued fraction. In the example of the precedinp^ article we divide 10 1 by 45 with a quotient 2 and a remainder 11 ; we then divide 45 by II with a quotient 4 and a remainder i. And this beinp; iden- tical with the operation for finding:: the G.C.M. of loi and 45, we deduce the following rule : Proceed as in findin<:^ the G.C.M. of the numerator and denominator of the given fraction ; the quotients taken in order form the denominators of the terms of the continued fraction. Ex. 163. To convert f|^ into a continued fraction. Proceeding to find the G.C.M. we obtain the quotients i, 2, 3, i, 6, i, 2, 2 in order. 2 .'. continued fraction is, i iiiiiiiii 1+2+3+1+6+1+2+2 2 472 681 I 54 209 3 7 47 6 2 5 2 I 00 ir . or more whicli is rrational |oi by 45 Je 45 by linpj iden- and 45, G.C.M. \tion ; the \ms of the I 9 5 I 3 6 2 00 — Ill — 4 I If we proceed to divide by the remainder o we get ^= oo , and the corresponding term of the continued fraction is i, which is zero. But as the process of finding the G.C.M. of any fraction must finally give a remainder o, the equivalent continued fraction must always be limited or rational. Hence any fraction may be converted into a rational continned fraction. 13S. If we take the values of one, two, three, four, &c., terms in the continued fraction of Ex. 163, we have, 1 I I_2 I I i_ - I I I 1_ 9 I I ' \ 1 ^ = |^,&c. 4-6 + 1+3 The quantities i, f, ^, -j^, |^, &c., are successively closer approximations to the value of the original fraction. They are consequently called convergents to the fraction |-|^. Thus the successive differences are : (I) i-W=W; (2)i-W=-*; (3) tV-W = M; (4) i^F-W=-M nearly; (5) U -|if = ^ nearly ; &c., &c., &c. We thus see that the 5th convergent differs from the origi- nal fi action by only ggf^o or -rrkw^ nearly. We see moreover that the odd convergents are too great and the even ones too small, so that the successive convergents are alternately too great and too small, the true value of the fraction always lying between those of any two consecutive convergents. To find the convergents. I I I I a + l)-\-c-\-d-\- &c. be the conveying fraction. Then ist conv. =-. For the second convergent we must 139- Let put a 4- i for a in the first; for the third convergent we must h put 6 4-- for b in the second ; &c., &c. c 112 We thus get, 1st convergent = - which denote by a 2nd "+* ba-+ I 3rd (( '-J ^^•+1 N, .b+i a{b + - which denote by N 3 4th i« We thus see that every convergent after the second is form- ed from the two proceeding convergents according to a fixed law, which may be stated as follows : Calling a, b, c, d, &c., partial quotients, the numerator of the 7j*'' convergent is formed by multiplying the number of the [n - i)*'' convergent by the tr^ partial quotient and adding the numerator of the (» -a)"^ convergent. The denominator of the «"* convergent is formed from the denominators of the (n- if^ and (« — 2)"** convergents in a precisely similar manner. The operation may be carried out as in the following ex- ample. Ex. 164. Find the convergents to the fraction, ^^. The partial quotients are 2, i, 3, 2, 4. Assume ^ as the first convergent ; then ^ is the second con- vergent. I 4 9 40 ^^132 4 I 2 3 II 25 III ) + a, ^3 Da" 2 C. is form- a fixed rf, &c., formed by the («-2)",1 formed ^^ergents /ing ex- :>nd con- — 113 — Write these two convergents in order and the rein.iining partial quotients in a row following them. Then starting with the partial quotient i as a multiplier, iXi (the numerator of ^) + o(the numerator of ^)= £, which write above for the third numerator. 1X2 (denominator of |^) + i (denominator of ^) = 3, which write below for the denominator of the third convergent. Next starting with 3, 3X1 + 1=4 for numerator, and 3X3 + 2=11 for denominator, &c. We thus find the convergents to be ^, ^, ^, ^ and finally the fraction itself -j^. Or the working may be arranged as in the margin, where the various steps are readily made out without any additional explana- tion. o I I 4 9 40 I 3 2 4 I 2 3 II 25 III Ex. 165. To find approximate values for 3.14159. Taiii :i plane surface, as in the margin, the whole number of objects in any square l)lock will be the square of the number upon the siMe. If a number of balls be piled in the form of a pyramid with a square base, each layer contains the square of the number of halls forming its side, and the sides of two con- secutive layers differ by unity. Hence the balls in the layers give the series of square numbers begining at the top where there is but one ball ; and the whole number of balls is the sum ot the square numbers from 1 to Ji^, n being the number of Jmlls on the side of the basal layer. Ex. 169. How many balls are in an unfinished square pyra- midal pile, the basal row having 22 and the top row 14 ? If the pile were complete there would be 2a(22H-l)(2X22 + l) _ 2 2X23X45 ■ ' ^ 6 ~ 6 • But the number required to finish the pyramid is * * * * * * * * * * * * * * * * * * * * * * * * * * -+ 1 — 117 — i3(i3H-i)(2Xii-|-i) _ 13X14x27 '. wholenumberinthepile = ^^^^3X45 13X14X27 ^^^^ :he right bers to « rs to the \rt. 93- * I * 4 * * * * * * * 16 luare pyia- )p row 14 ? SERIES OF TRIANGULAR NUMBERS. 2 143. If objects be arranged in equilateral triangles upon a plane surface, the number required to form a complete triangle, as in the mar- 1*1 gin is called a triangular number. With I object upon a side we have i as the first triangular number. With 2 ob- jects upon a side it requires 3 to com- plete the triangle ; there being one row with one in it and a second row with two. With 3 upon a side we have 3 rows, of I, 2 and 3 objects respectively; i.e., 6 in all. With 4 upon a side we have four rows of i, 2, 3 and 4 objects, or 10 in all, &c. Hence the series of triangular num- bers is I, 3, 6, 10, 15, 21, &c. » * * * * ¥e * * 6 10 « « « The numbers are evidently the successive sums of the series of natural numbers beginning at unity. Thus, 1 = 1,3 = 1 + 2,6 = 1 + 2 + 3,10=14-2 + 3 + 4 15 = 1 + 2+3 + 4 + 5, &c. 144. To find the sum of « terms of the triangular numbers. Let Hn denote the sum of n terms of the series of natural numbers, 2!n^ that of the series of square numbers, and 2t the sum of n terms of the series of triangular numbers. Then, 2w = i +2 + 3 + 4 ... .«, 2'«2 = 1 + 4 + 9+ 16 n"^, In-^ In^ = 2 + 6 + 12 + 20+ ... .{n^+n). = 2(1+3 + 6 + 10+ '^'* ) — ii8 = 2iV. . Vi i/v I V2\ I'Kw+i) , n(n-\-i)(2n-\-i) . _^ „ .-. 2.t = ^{2n-\-2^) = ^-~ ' -{■ —^ — '—-^ ■ — ' Arts. 93&142. , _n{n-\-i){n-\-2) If a number of balls be piled in a triangular pyramid, the numbers in the successive layers will be the series of tri- angular numbers, and the whole number of balls in the pile, commencing at one upon the top, will be the sum of the first n triangular numbers, n being the number of balls in a side of the basal layer. Ex. 170. How many balls in a complete triangular pyramid, the basal layer containing 10 upon a side. Here w=: 10, and, , ,• - i'/ = 10. II . 12 = 220. INDETERMINATE COEFFICIENTS. 145. The truth of the statement that = I -{-x+x^ + .... ad inf. I -X is not limited to any particular value of x, but holds for all values, arithmetically if .«: is less than one, and algebraically if X is any quantity whatever. In other words, the expression is an identity, and must, therefore, be true, quite independently of any particular values given to the symbols employed. 146. Proposition. If we have the identity A +Bx-\-Cx^ . . . .=a + bx-\-cx'^+ .... Where A, B, C .... a, b, c\ . . . are constant coefficients, and .r is variable, then, A =a, 3 = 5, C=c,&c., i.e., the co- efficient of any particular power of x upon one side of the identity is equal to the coefficient of the same power of x upon the other side. For, — 119 — A -a = (b- B)r + (c - C)x^ + But the second member chanp;es value as x changes its value, v^'hile the first member is constant. Hence there cannot be equality unless each member is equal to zero. .'. ^ -a = o or A=a, and by rejecting A and a as being equal and dividing by x we obtain in like manner B -b = o, or B =b, &c The coefficients A, a, B, b, &c., are called indeterminate or undetermined coefficients, and the proposition now proved states the principle of indeterminate coefficients. The principle of indeterminate coefficients is one of the most prolific in algebraic analysis. Some of its simpler appli- cation will be illustrated by a few examples. Ex. 171. To expand ing powers of x. l-irX into a series according to ascend- Put ^"'"^ ^a+bx+cx'^+dx^-{- . . . . (l-;ir)2 then i-^x={i-2X+x'^){a + b->tcx'^ -\-dx^-\- = a +b x^c x'^^-d -2a -2b ~2C + a + h and equating coefficients of like powers of x, a = i ; h — 2a=i .'. 6 = 3, c-2b-^a = o .'. c= 2b -a = $, d-2c-\-d = o .'. d = 2c-d = y, ■i\ &c., &c. .•■ ^^^ ^ = i+^x+Kx^+yx^-\- . . . . / (1-^)2 ^ . V ■■ pompare this result with Ex. 22. Ex.r 172. To expand the square root of i i-x+x^. Put ■/i+x-\-x^= a-\-bx-\-cx^ +dx^+ .... Squaring, i+x-\-x'^ =a^ + 2adx + 2ac x^ + 2ad zbc I20 — Equating coefficients, a^ = i .'. a = i 2ac-\-d^ = 1 .'. C s= =-# 2a 8 2ad-\-2dc=o .'. d— _ _^ ^ __3^ i6 .-. |/i +^4-jr'» = I + |;r+|jr2 - j^j;< . . . . i6 Ex. 173. What relation must exist among the quantities p, g, y, s in order that x^+px-k-q and ^ + rx^s may have a common factor. Let the common factor be x+a, then the expressions may be written \x-{-a){x+^\ and (;r + a)(;r4-l), a a since the last terms in the products will evidently be q and s as they should be. Then we must have, a+l-p, a-|-- = y. a a .'. c^ — ap— — q, a^-ar= -s. And eliminating a^ and a by determinants. 9 I s I «= \p I r 1 pq and a' = r s I r p — r p~r .'. (p-r){qr-ps) = iq-s)^, is the necessary relation. tides/, qy ay have a sions may be q and