]\gt ; [ 151 ] VI .
On the meters of Plane Cubic .
J. J. WALKER , M. A. , F.R. Received June 16 , \mdash ; Read June 16 , 1887 .
Revised February 9 , 1888 .
[ PLATES 6-8 .
] I. ABSTRACT .
1 .
THE object of this Memoir is to develop relations which subsist between a cubic ( u ) and the complex of lines , in its plane , which are the polars with respect to it of the points on any transversal ( L ) .
This complex becomes the system of NEWTONIAN DIAMETERS of the cubic , when the points on the plane are projected on a second plane parallel to that contaimng the vertex of projection and the line .
development involves frequent reference to envelope of the complex in question , the conic , ( 1 ) which , in analogy with the " " pole\ldquo ; of a line in the theory of conics , I propose to call the " " POLOID\ldquo ; of the cubic and the line ; and , in particular , when the line is at infinity , the " " CENTROID\ldquo ; of 2 .
HESSE* appears to have used the equation to in the theory of the ternary cubic form\mdash ; but without any recognition of its geometrical significance\mdash ; to obtain the equation to the cubic in " " line-coordinates:\ldquo ; viz. , in the form of the resultant of the system with and this resultant will plainly be , to a factor , the equation to itself in coordinates , with substituted for respectively .
*'CRELLL , Journ. Math vol. 41 , p. 285 . .
, under dato January , 1850 .
28 .
688 .3 .
In the first edition of the ' Higher Plane Curves ' ( 1852 ) SALMON , under the subject " " Poles and Polars\ldquo ; ( of cubics ) , showed that is the envelope of the polar lines of points on , touching each tangent to at the points in tlJe point harmonic conjugate to its contact with relatively to its intersections with the other two , and proposed for the locus of the designation " " Polar conic of the line\ldquo ; , for which I have yentured to suggest , and use , the shorter name " " Poloid\ldquo ; of ( and ) L. In CAYLEY'S Memoir " " On Curves of the Third Order ' Phil. Trans 1857 , the additional property is given of the polar lines of points on all through a point ( viz , the intersection of the two right lines into which , " " the lineo-polar envelope of the line then breaks up ) on the Hessian of , when Ljoins corresponding points on that curve .
Beside these I have not been able to find any notice of the conic 4 .
If through any point in its plane chords are drawn meeting the cubic in , and a point is taken on the chord determined by the relation the locus of is a line , according to a theorem of COTES'S , communicated by his friend , Dr. ROBERT SMITH , Master of Trinity Colflege , Cambridge , to MACLAURIN , after COTES'S lamented death , and proved by MACLAURIN* as a case of a more general theorem which presented itself to his mind wheI ) " " meditating on this communication For shortness I have called the point the " " CoTEs-point\ldquo ; on the chords through , the locus of which is now well to be the polar line of the point relative to the cubic 5 .
If ibes a line , the -points of the polar lines of the points on regarded as chords of u\mdash ; relatively to their intersections with , may be considered .
The locus of the -points of this complex of lines is shown in the sequel to be a nodal cubic ( 37 ) and ( 40 ) , which covariant of and the line , I propose to call their " " 6 .
Considering 33-37 more generally the locus of -points on chords of subject to the condition of touching , the result comes out as a concomitant breaking up into two factors , one the cubic just referred to , the other being the equation to * Lin .
Geom. Prop. Gen. Theor .
p. 24 , ed. 1748 .
The Theorem is ot given in the ' Harmonia Mcnsurarum , ' as sometimes erroneously stated , with the subjects of which treatise it has no connexion .
MR. J. J. WALKER ON THE DIAMETERS OF A PLANE CUBIC .
the three tangents to at the points in which it is met by , now in a general form .
In the Memoir referred to Phil. ] obtained , with his peculiar skill , the equation of the ents in question for HESSE'S canonical form of ?
( 2 ) finding for the " " satellite line\ldquo ; of } line in which the tangents meet , the equation that , furthel on ( S 36 ) will be shown to verify the at which I have arriyed ; and in terms of which ( if it be taken a iant of ?
and L ) and of the equation of the nodal cubic 1 ' ) be 7 .
Whereas through any point in the plane of the cubic th line there be drawn in general ( SS 40-43 ) two chords it as their CoTEi3-point relatively to those in which they meet , viz. , those determined by the of the point ; if , however , that point be taken on the two chords coincide , a complex of double chords is obtained , which the polars with regard to 42 ) of the -points of the ] ines of the points of , and shown 58-60 ) to have as their envelope a tricuspidal quartic ( 81 ) the equation of which cannot be found } ) licitly except for assigned forms of The point in which a double chord meets being its -point , that in which it touches its envelope is shown to be harmonic conjugate to the former with respect to the intersections of the chord with the line and with the conic again ; and the two intersections of the double chord with thereby d ated 59 ) .
8 .
Again , considering ( S 61 ) the points of a line having its pole on , the chords of which those are -points constitute two groups : vi , one pencil pole of the polar line , the other as its envelope a conic 88 ) touching the line as well as the polar line in question ( in virtue of its being the line of the lattel complex through its own CoTEs-point ) and the double chord through its ) oint of contact with , which is both a ray of the pencil and line of the complex .
This system of conics ( S62 ) has as its envelope the sides of the quft formed by the transversal and the tangents to the cubic at the points in which meets it .
9 .
A question of some interest is considered 38-39 ) : what , , of the complex of polar lines of points on the transversal are , in the sense of their section being the -point on either ding the ngents to at the points in which it is met by the transversal , each of which is conjugate to tlpe distinct conjugate polar lines of points on are the two to poloid ( s ) of from the pole of that line with respect to the poloid .
10 .
As the chord of contact with the poloid of the two ngents through its pole forlls a coincident pair of\ldquo ; double chords\ldquo ; ( of the cubic ) ; and these three lines , , the two tangents and their chord of , form a triad of lines of reference by means of the properties of the complexes of lines here considered may be deduced with far gl.e , atel .
facility the use of the canonlcal form ( 2 ) .
The two tangents to from the pole of with respect to are the nodal tangents of the Cotesian ( S 50 ) , the line being its inflexional axis ; a circumstance which ains the unique character of this triad of lines , and marks them out as the best system of lines of reference for the discussion of properties connected with this concomitant of the primitive cubic ( SS 48-62 ) .
11 .
But , these tangents to , the poloid of , are only real when cuts in real points ( which is shown to occur only when it meets the cubic in but a single real point ) , three other triads of lines exist , of which one at least is always real , convenient as lines of reference in many of the questions which arise : , the sides of one of the three triangles whose corners are the pole of , with respect to ; one of the three points in which meets ; and the pole of the connector of these two points .
Each of these being self-conjugate triang ] in respect of , the equation of that conic is reduced to a trinomial form ; and since the cuspidal tangents of the envelope ) all pass through the pole of , the equation of that curve is of a comparatively simple character for one of these triangles of reference .
In Plate 6 , ABC is one of these triangles ( SS 63-76 ) .
12 .
If the line touches it touches its poloid also ; and consequently other lines of reference have to be looked for .
These are found in the line , the tangent to at the point where cuts it itself tangent also , will remembered ) , and the chord of contact of these lines with In this case the cubic degenerates into the line , and a conic having double contact with ; while the envelope also degenerates into a conic having double with and the Cotesian conic at the same points ( SS 77-81 ) .
13 .
When becomes the line at infinity the pencil of chords through any point on it is to be replaced by a system parallel to a given line , and the polar line by the diameter which is the locus of the mean point on any chord of the system relatively to its intersections with bhe cubic .
The envelope of these diameters I have called the " " Centroid\ldquo ; of the cubic , from its evident analogy with the centre of conics , apprehending no confusion in such a connexion with the sense of a " " mass-centre , \ldquo ; which it sometimes bears .
The consideration of the Centroid and associated curves occupies the concluding part of this Memoir .
14 .
The method of treatment of the discussions , an abstract of which has just been given , is uniformly analytical trilinear coordinates being employed .
The results will be found to be arrived at without much difficulty , or tedious calculation , considering the great generality of most of them .
With a view to simplifying three important discusMR .
J. J. WALKER ON THE DIAMETERS OF A PLANE CUBIC .
sions as much as possible , a preliminary investigation of a form into which the result of substituting for the yariables in a ternary quadratic form the expressions being ternary form three constants , introduced ; and the special cases of its application in the sequel are considered .
15 .
The notation employed throughout is as follows:\mdash ; The right line , considered generally , is written : The particular tratlsversal considered in connexion with a cubic is always written : ( iii .
) The cubic being , in point coordinates , is written in line coordinates\mdash ; or its reciprocal is\mdash ; . . . .
: ( iv .
) The poloid of with respect to the cubic in coordinates is represented by as in S 1 , or ( S19 ) .
, and in line coordinates\mdash ; or its reciprocal\mdash ; by , where Any special point on regarded as the " " pole\ldquo ; of a pencil of chords of the cubic is marked ( x'y'z ' ) , and an .
special point on the polar line of x'y'z ' is marked " " chord\ldquo ; being a line , or transyersal , considered particularly in connexion with its intersections with the cubic II .
PRELIMINARY .
16 .
The intersections of a line , with a cnve may be studied through the equation , ( 3 ) where , x'y'z ' being a ) given point on the line , cmd defined by , ( 4 ) if , for tness , epresent the sines of the angles of the triangle of reference , is to the length of segment of the line between the point x'y'z ' and any one of the / ) points in which it meets the curve ( ' London Math. Soc. Proc vol. 9 , .
Thus the sum of the of those segments is equal to ; while , if is other point on the same line , , and similarly equal to the segment veen x and it , , , . . .
( 5 ) But if is the -point on the line in respect to x'y'z ' and the curve ; , ( 6 ) , the locus of the polar line of .
If now , instead of considering as yariable and as fixed , are equal as ally point on , or , the envelope of the line ( 6 ) will obviously be the same as its in line coordinates , regarded as a curve of order in poirlt coordinates , if were a quartic , ( 3 ) having been written in the form its envelope at once be written down as ' ; viz. , the poloid of the line and the quartic is another quartic .
I have made the foregoing remark to draw attention to the perfect generality of the theory of the MR. J. J. WALKLR THE OF A CUBIC .
poloid , and , at the same time , to show it is of equal or higher degree than for values of exceeding 3 .
19 .
In the case , then , alone contemplated in this Memoir , of being a cubic , and ( 6 ) being , tlJerefore , otherwise written the envelope takes the form ( 1 ) mn as given by SALMON Higher Plane Curyes , edition , here replacing If written in the normal form of a conic the coefficients are the contravariant conics of the triad of conics , ' 1 ' ( 7 ) taken singly and in pairs , , the cubic being written as in the 'Higher Plane Curves , ' only with the substitution of for exyz , ( 8 ) and the poloid of and being , ( 9 ) then 20 .
The inval.iants of the equation ( 3 ) , which for , is , . . . . .
( 11 ) and , for one of the conics , .
, or .
, , . . . . .
( 12 ) give the fundamental invariants of and .
, in a very succinct form , involving regarded as rameters connected by the equation ; thus , if then ( 11 ) being replaced in the operator , , by . . .
; . . . . .
( 13 ) ' . . .
; . . . .
( 14 ) ( ' London Math. Soc. Proc vol. 9 , p. 232 .
) Also ( 1 ) , . . . . . . . .
( 15 ) , ( 16 ) if .
is the standard form of the condition that shall touch 21 .
The above forms are very convenient for the comparison of related concomitants essential to the objects of this Memoir .
Thus , the condition that , should touch . . . .
, ( 17 ) multiplied by ' is equal to Now , since , whatever integers may be , * It is to be observed that if is of order , then is of order \mdash ; and that if \mdash ; of order of order , then ; an at if being and so on : thus ; 4.3 MR. J. J. WALKER ON THE DIAMETERS OF A CUBIC .
or ( 15 ) .
( 18 ) But 3 D ; . . . .
( 19 ) and or By substitution from ( 19 ) ( 20 ) in whence ( 16 ) ( 17 ) , ( 21 ) i.e. , the condition that shall touch , is equal to onefourth of theft for the cubic , with changed sign . . . . . . . . . . .
( i ) 22 .
Considering next the discriminant of ; viz. ( 9 ) , if , for shortness ( 7 ) , ( 23 ) ( 24 ) then , ; . . . . .
( 25 ) * By means of these expressions the equation of the poloid ( s ) may be thrown into a form which exhibits it explicitly as the envelope of the polar line of points in : viz. , by ( 9 ) , ( 13 ) , ( 14 ) , ( 23 ) , . . . .
( 8 bis ) Of the right lines in this form , ( i ) is the polar line of the point x'y'z ' in ; ( ii ) or ( is the Newtonian diameter of chords of parallel to ( shown in fig. 1 touching the poloid at ) or is the chord of contact with ( s ) of ( i ) , ( ii).\mdash ; ( April , 1888 .
) .
( 26 ) Hence ) , , ( 27 ) ; viz. , is the colldition that should be cut in involutio by which condition , multiplied by , has been shown ( ' London Math. Soc. Proc vol. p. 233 ) to be Thus it is proved ( 27 ) that four times the discrir , of equal to the square of the Caleycm of ? . . . . . . . . .
( ii ) 23 .
If the line joins corresponding points on the Hessiau of its coefficients satisfy the Cayleyan ; hence its poloid breaks up into two ight lines through the intersection of which the tangents of all pass ; and the locus of this intersection , as the line varies in position , is the Hessian itself This is the property mentioned in the Introduction ( S 3 ) as proved by CAYLEY , : Phil. Trans 1857 , p. 432 .
24 .
A theorem on the result of certain substitutions , which will be of use in subsequent investigations , may be conveniently considered before entering upon them : If is any ternary form of order , and a quadric\mdash ; say , . . . . .
( 2s ) .
J. J. WALKER ON IETERS OF A PLANE CUBIC .
and if in for are respectively \ldquo ; the result may be thrown , if , into form . .
( 29 ) The proof of this theorem depends on the identities , ( 31 ) with similar values for But ( 28 ) MR. J. J. ON THE ] RS OF A CUBIC .
; the substitution of the values ( 30 ) , ( 31 ) of , , , .
, in which , with tho addition and subtraction of the terms , wherein 2 lx , , gives an expression identically equal to the form ( 29 ) .
25 .
The cases of the application of this theorem which occur are : When ( S 28 ) is a cnbic , and is the polar line of a whose coordinates are .
, ( x'y'z ' ) being also a ; so that . .
: in which ' and the quadratic functions in which multiply and respectively become . .
, . .
, and rpesectively , where Now , otherwise , the substitutions may be made in the conics , .
of the other of the ) line , MR. J. J. WALKER ON THE DIAMETERS OF A PLANE CUBIC .
and then the multiplier of 3 ( present value of in ( 29 ) ) becolles , for ; or , identically , that multiplier becomes for respectively .
Thus the result of the substitutions becomes ; ( 32 ) whereas , by making the substitutions in the first way , the result would be .
Identifying these two ; so that the polar of x'y'z ' with respect to the poloid of may be written 26 .
In the second case which occurs of similar substitutions .
the doubles of the coefficients of ; and no longer vanishes .
The result ( 29 ) , S 24 , thus becomes . . . . . . .
( 34 ) It is to be observed that the values of .
are to be obtained by substituting .
for a. after differentiation ; thus , generally , in the present case , or 27 .
Lastly , the case which will occur at S38 of the substitutions , , .
for in differs from the first case only in the differential coefficients of being multiplied by instead of , so that ( 29 ) becomes simply equal to ; i.e. , \mdash ; us with x'y'z ' for III .
GENERAL FORMULB AND EQUATIONS .
28 .
The polar line of a point ( x'y'z ' ) on , or , meets it in a second point , the of which are : : ; and if the coordinates of this latter point are substituted for in the quadric forms , , of the polar line oi this point will be expressed in terms of the coordinates of the point .
MR. J. J. WALKLR ON THE OF A PLAN CUBIC .
Refell.ing to the general form for the result of such substitutions , ( 29 ) , in the case it is ( 32 ) , S25 , , ( 35 ) the form of which shows that the polar with respect to the poloid on cuts the polar with respect to of point in its 29 .
The relation just proved gives at once the coordinates of that -point in the form And , plainly , the -point of the polar line on is the pole ( with respect to ( s ) the poloid of L ) of tlmt chord of the pencil through the point , uhich through the contact )of the polar line of 30 .
From considerations founded on the relation just established the locus of the CoTEs-point of the polar line of a point ( x'y'z ' ) on may be at once obtained in a general form , viz. , by the elimination of among or or From the last two : : and the substitution of these values in the first gives the locus of in the ( 34 ) S26 , of six times , if 166 MR. J. J. WALKER ON DIAMETERS OF A PLANHi CUBIC .
, ( 37 ) where , as above ( 7 ) , , are the contravariants of taken singly and in pairs , these being the coefficients of the poloid , viz. , ; and are the coefficients of its reciprocal , viz. , . .
31 .
The coefficient of in the above expression admits of an important transformation , , it may be written . . . . .
this ( 7 ] ) , S5 is identically equal to , . . . . .
( 38 ) where is the Pippian , or Cayleyan , of the cubic Also ( 55 ) . .
, ( 39 ) MR. J. J. ON OF A PLANE CUBIC .
, . .
, . .
being the second differential coefficients with respect to of the reciprocal ( v ) of . .
; while , otherwise , ( 21 ) .
With these substitutions the equation to the locus of the -points on the polal lines of points in , or , takes the form a cubic , enerating into a conic when the line touches the cubic , which will be traced further on ( SS50-53 ) by referring it to the line and the two tangents to the points .
But previously it will be of interest to show the signifi- cance of the part \mdash ; 2 ( 41 ) in the equation to ; and to add a few remarks on the relations ( 38 ) ( 39 ) .
32 .
It is not easy to verify these relations by means of the invariants of ( 12 ) S20 , because the variables which enter into them are perfectly general\mdash ; except satisfymg constant .
They are verifiable , the former with ] ight , the latter with moderate labour by means of the canonical form of the equation to the cubic ; but much more readily by means of the simultaneous forms of and to which form of and its concomitant are reducible , referred to in S10 .
The verification is therefore deferred until the reduction to those forms is explained in the sequel , and will be found in the paragraphs cited .
33 .
The tangent to at any one of the three points in which it is met by the transversal , being the polar line of that point , also totlches the poloid , the point of contact with its CoTEs-point\mdash ; viz. , it is that determined on it by the polar with respect to of the point in which it meets ; or , otherwise , as the point in which it is cut by the coincident tangent .
The general equation to the three tangents , at the points , may be obtained without the difficulties which would attend the direct hitherto unattempted , at least successfully\mdash ; through the property of their touching the poloid also , as follows : the point of contact of one of the tangents being ( x'y'z ' ) , its equation is or , being a cubic , .
( 42 ) J. J. THE DIA ) OF A PLAN CUBIC .
But in virtue of its also , if is any other point on it , the or tangential of ; ' . . . . . . . . . .
The eliminant of these two with in the known form , ( 44 ) is , if 34 .
For gh \mdash ; ( Discrt .
of s ) .
Hence . . .
( 45 ) Observing now that . . .
, it appears at once from S24 that\mdash ; substituting for MR. J. J. WALKER ON DIAMETERS OF A PLANE CUBIC .
so that . . . . . . . . .
the negative term in which being the result of substituting for . .
in is ( S80 ) ; . . . . .
( 46 ) consequently , ( 45 ) ( 46 ) , sixteen times the eliminnnt ( 44 ) is . . . . . . . . .
( 47 ) As was mentioned , ( 38 ) , S31 , the terms multiplied by within in ( 47 ) are equal to so that the eliminant is ] , after division by 36 , or ,35 .
The presence of the ctor v in this result is CCOUnted for by the fact that the solution really that of the more general question : " " to find the locus of the -points of ffi of ?
( elatively to their intersections with L ) which touch \ldquo ; ; discarding , therefore , this factor , the equation of the tangents to at the points .
( 48 ) now to the form ( 40 ) of , in which . . .
, the equation of the three tangents to at the points is found in its standard form ; viz. , , ( 49 ) wherein are to be replaced in and its second differential coefficients by 36 .
It will be satisfactory to verify the general expression for the satellite chord of by applying it to the canonical form of , for which CAYLEY has obtained the form lefelred to in S6 .
For exyz , ; whence , Again , MR. J. J. WALKER ON THE DIAMh'TERS OF A PLANE CUBIC .
, ; whence , ( 51 ) From ( 50 ) ( 51 ) there results , as the equation of the satellite chord of ; . . . .
( 52 ) which , if , and be replaced by , agrees exactly with the form given in the ' Memoir o11 Curves of the Third Ordcr\ldquo ; ( ' Phil. Trans 1857 , p. 439 ) .
37 .
rfhe values of the second differential coefficients of and used in the preceding verification may be calculated from the formulae ( 10 ) given , S 19 .
all the coefficients of having suffixes vanish , .
for the present form of are found ; and thence .
obtained .
Next , expressed in terms of as line coordinates , is determined as viz. , therefrom it is found that Finally , the second differential coefficients of may be found .
And the value of so found affords an independent verification of the relation , S21 , when in and have been replaced by I may remark , in conclusion , that I first obtained the general form of by considering the question mentioned in S35 , determining in the resulting form ( 47 ) which of the two factors represented , which , by examination of the special forms when the line and the tangents to at the points , were the lines of reference .
ubsequently , I observed the method of arriving at the general form of given SS 30 , 31 , independently ; which completed the general proof in as simple a manner as could be expected .
38 .
The triad of tangents reappear as a part of the complete solution of a terest iexion with the of this Memoir : what , if of on are , in sense of having a common ?
line of the point in the point : ( 52 ) ondil i plainly is that line of ( x ) ( z ) should pass through the when for in .
the values just given in terms of are nbslilutc The result is at once by means of the general theorem , ( and 27 ) ; it is simply with ] points plainly coincides with ' thus ngcnt to ?
at tbose points is a double conjugate line , which ccounts itlle factor in the result above .
39 .
The points determine unique of distinct or proper polar ; viz. , these are the it is by ; which will be real ) when mcets in real points .
ents to .
chord of contact form , as before remarked , unique of lines , to which ] the cubic ?
and its systeln may be convcniently referred , in questions in volving , and certain other concomitants of and particularly , will } ) .
in the sequel .
Plainly the equa , tions thus referred will only be as two coordinates , however .
40 .
in the plane of the cubic and the transversal two ords of can , in , be drawn having it as their CoTEs-point , relatively to their intersections with , the collnectors of with the points . .
in which ] its meets If is itny othel point on one of these chords , then the coordinates of the point it meets MR. J. J. WALKER ON DIAM [ OF A PLANE CUBIC .
will be proportional to or and the substitution of these coordinates in gives the equation of the two chords in question , in the form wherein stands for for The equation of that one of the pair which cuts in the point ( x'y'z ' ) being that of the other will be either , ( 53 ) or one of the two analogous forms obtained by interchanging or 41 .
The reciprocal of the equation ( 52 ) being , to a factor , if should be a point on the poloid , the two chords having it as their COTESpoint will coincide\mdash ; as would otherwise be inferred from the consideration that any point on might be regarded as the point of intersection of two coincident en ts \mdash ; or become a " " double chord and its equation will be either ( 54 ) or one of two allalogous forms resulting from the interchange of or Otherwise , the equation of the double chord through , a point on is .
( 55 ) 42 .
But considering a point on , the double chord through it is , plainly , that connecting it with the point in which its polar line touches the envelope , viz. , this is the polar with respect to of the on that polar-line .
Its equation , thus considered , is therefore with the coordinates of the -point referl.ed to substituted for in ; those dinates being ( 36 ) , S20 , ou ' on ' ' The resulting equation to the double chord is , therefore , .
( 56 ) 43 .
This being cubic in , and of five dimensions in the coefficients of , its envelope , obtained as the condition that shall touch the ternary cubic in x'y'z ' , will be a curve of order 4 in , of order 6 in lmn , and of order* 8 in the of envelope , as well as that of the complex of chords of , connected with the polar line of a point on by their -points on it , but not passing through its pole , will be more conveniently considered by means of special forms of and to which every cubic be reduced .
*This would appear to be 20 , but another form of the equation of the double-chord , into which the coefficients of enter only in the second degree , may be obtained from that of ( 8 bis ) given in the Note to S 22 .
Combining ( i ) and ( iii ) of that Note , the coordinates of the point of contact of the polar line of ( x'y'z ' ) with the poloid are at once given as proportional to respectively ; hence , the line joining this point with is But , if stand for . .
; for respectively , . . .
; whilc , since m Substituting in ( i ) the equation of the double-chord ( DE , fig. 1 ) becomes divisible by , and is . . . . . .
( 66 bis ) Note added April , MR. J. J. WALKER ON DIAMETERS OF A PLAN CUBIC .
44 .
The question of the angJ e( ) between the polar line point on and the " " double\ldquo ; chord of the pencil through ) point is of some interest , as a genera ] sation of the question of the angle which a Newtonian ' diameter\ldquo ; with its ' ' ordinates The polar line being and the double chord ( 56 ) what may be called the " " direction coordinates\ldquo ; of the two lines are ( the angles of the triangle of reference being ) and , using as in S20 ; .
as in S30 , , , .
The angle between two lines whose direction coordinates are is given , generally , by , ( 57 ) and in the particular case now under consideration A ( 58 ) 45 .
If is the point of contact with its envelope , the poloid , of the polar line of a point x'y'z ' on or from the forms of the equation of that polar line it is evident ' ence at once iven in terms of : : , ( 59 ) . .
being the coefficients of the reciprocal of Conversely , if are to be expressed in terms of they are the coordinates of the point common to the two conics and the line 46 .
But the point is readily determined by the double line of the pencil of chords through that point , having their -points on the tangent to the poloid at , which has been shown to be ( 54 ) At the intersection of this and giving : : or either of two other analogous sets of values .
47 .
Thus , when is taken as , and of the coefficients .
of th tangential equations of the poloid , all vanish except ; MR. J. J. ON DIAM l OF A PLAN CUBIC .
giving for the point of contact of the polar line of ( x'y'o ) : : : ; and conversely IV .
SPECIAL FoRblS 0F CUBIC AND PoLOlD .
48 .
now treat some of t , he questions , hitherto considered without reference to any particular form of , in more detail , by means of special forms in which the line is used as one of the lines of reference--.say ; . . . . .
( 63 ) and for the two other lines of reference I take 1st , the tangents to the poloid from the pole of , as The conic is now ( 10 ) S being dropped\mdash ; reduced to with the conditions , . . . .
, . . . . .
( ii .
) \mdash ; , . . . .
( iii .
) \mdash ; . . . . . .
From the first and second of these , whence since is excluded by the form of Hence , from the second and from the third and fourth MR. J. J. WALKER ON THE DIAIIETERS OF A PLANE CUBIa since is excluded by the form of , now reduced to , . . . . .
, . .
( 64 ) is reduced to the terms . .
, ( 65 ) 49 .
The line meets in the points being the imaginary cube roots of \mdash ; 1 , so that , and the tangents to at these points are ( 66 ) the common equation to the three being , by actual multiplication , \mdash ; 8 , . . . . .
( 67 ) ; restoring lapsed factors , .
( 68 ) Further , the same equation may be thrown into the form 50 .
The tangential equation of is now simply where MR. J. J. WALKER ON ] DIAMETERS OF A PLANE CUBIC .
17 !
The general value ( 37 ) of , the Cotesian of , now reduces to exyz ) - ( ) , . . . . .
a cubic having a node at the pole of now with respect to , the tangents to being also the nodal tangents of the Cotesian .
The comparison of the equations ( 69 ) ( 70 ) of and verifies the relation , ( 48 ) S37 , since , in the present form ( 65 ) of , the Cayleyan , for the values of the line coordinates , is simply . . . . . . .
( 71 ) 52 .
Eliminating and successively between there result showing that has triple contact with ; viz. , when the loop of is real it touches at the point ; or at the poin determined by the real lines , or any pair of them , Now , making in and , the points in which meets these cubics are determined by the real one being so that tl , mcctors of the pole of ( with respect to s ) with this point and the point conjugates relatively to the tangents to from 53 .
The polar line of , a point on , meets the Cotesian in three points point ) and two other points , while it touches the poloid in a point D. Now is also the -point on this polar line relatively to the point and cubic v : to prove this\mdash ; The of , ( 59 ) , , : , ( 72 ) ( 73 ) becornes , for those values of the coordinates of : : ' : . . .
( 74 ) , the -polar line of is ; ( 75 ) but ex'y'z . . . . . . . .
the -pclar line of , and , . . . . .
( 77 ) the -polar of the same point , cuts it in its -point , , .
Thus the three lines ( 75 ) ( 76 ) ( 77 ) meet in one point , and this is the U-COTES-point relatively to , and the V-COTES-point to of ( 76 ) ; whence or .as verified very exactly in figs. 2 , , i.e. , the -point on the polar line of point in is harmonic-c , onjugate to its point of contact with , poloid of and relatively MR. J. J. WALKE'R ON THE DIAMETERS OF A PLANE CUBIC .
to the other points in which it the ; and is thus discriminated from them , when all three are real .
If one only is redl , that of course is the COTESpoint .
54 .
The general form given above ( 49 ) , S 35 , for the satellite chord of the line may be verified from the form it takes for the present one of , the reciprocal of which\mdash ; as far as the terms giving those which do not vanish in its second differeI ) tial coefficients for \mdash ; is ; . . .
( 78 ) viz , , for those values of ; while the reciprocal of being now Also , .
With these values of the second diflerential coefficients , which do not vanish , for the present forms ( 64 ) ( 65 ) the satellite chord of , or , is multiplied by , or , identically , which agrees with the form in .
55 .
What is , , a verification of the relation ( 39 ) ; but to it more clear , observing that the only one of the ents A ' of the ) of does not disappear from the slnlster of ( 39 ) , in the term , and adding to this , that sinister becomes Again , the dexter is being now ez or , identically , also Lastly , the only term in the sinister of ( 38 ) , $ 51 , .
It has been shown that and meet the transyersal in the same three points ( S52 ) , of which one is real when the line meets its poloid in real points , the condition of the lines of reference , at present used , being real .
That this would be the case appeared from the fact that the discriminants of the cubic ( 12 ) and quadratic ( 13 ) S20 , whose roots are proportional to the intercepts of between a certain point on it and its intersections with and respectively , are of opposite signs .
57 .
The tangents to at the points in which it is met by are plainly obtained fxom those to at the same points , by changing into ; viz. , they are ( 66 ) MR. J. J. WALKER ON THE DIAMETERS OF A ] ] CUBIC .
eliminating between this equation and , the result is , the points in which is met by are points of inflexion on , as was manifest from the form of its equation .
When the node of is real , only one of these points of inflexion is real ; but if ) is acnodal , all three will be points of real inflexion .
Thus mucf ] is known of generalIy .
In Plates 7 , 8 , it is figured for two cases when is the line at infinity , as will be further described .
58 .
The coordinates of the -point on the polar line of x\fnof ; m ex'y'z may be found as those of its point of intersection with the -polar of ( x'y'O ) ; , they are : : : 2 : ' . . . . .
( 79 ) or , what is the same thing ( 36 ) , p. 165 , : . .
: .
The double chord ( 56 ) S42 , through is determined as the polar with respect to of that CoTEs-point ( 79 ) on the polar line of ( x'y'O ) .
Its equation is , therefore , Arranging this as a binary cubic in , . . . .
( 80 ) the discriminant will be the envelope of the double chord as the point ( x'y'O ) val.ies on the transversal , viz. , or , developed , 59 .
If the first derived of the cubic ( 80 ) , i.e. , ( 82 ) are combined , the coordinates of the point of contact of the double chord with its envelope will be obtained in terms of those of the point on , through which it is drawn , , eliminating between the equations ( 82 ) , the point of contact is determined by the line .
( 83 ) The section of the double chord with lies on the line , . . . . .
( 84 ) and its two intersections with the poloid on the lines , . . . . .
( 85 ) , . . . . .
( 86 ) the latter being its CoTEs-point , viz. , the point of contact with of the polar line of , S47 .
Now these four lines through , form a harmonic pencil , since , writing ( 84 ) ( 86 ) [ giving then ( 85 ) and ( 83 ) x'y'q ; , the later pair are harmonic-conjugate with respect to the former pair .
Thus it appears that the -point on double chord , and its intersection with the line -conjugate points with respect to its second ( or 'empty intersection with the , and its contact with its envelope , as stated , S7 .
This is shown in Plates 7 , 8 , in the case in which is the line at infinity , by being the mid-point of 60 .
Returning to the equation ( S1 ) of the quartic , if and be alternately elimmated from either MR. J. J. WALKER ON THE DIAMETERS OF A PLANE CUBIC .
18.5 , or by and there result showing that is a tricuspidal quartic , two of the cusps being imaginary when the line meets the cubic in one real point only , the real cusp being at the point : : on the line which is the cuspidal tangent , and is -conjugate to the connector of the -pole of and its real intersection with , relatirely to the tangents to from that pole .
The curve is shown in the figure , Plate 7 , in the case when is at infinity and the cubic has a single real asymptote .
61 .
Through any poin on the polar line of a point x'y'z ' in , two chords of pass which have the first as their CoTEs-point ; viz. , the chord which forms one of the pencil through x'y'z ' and the other ( 53 ) , S 40 the line , being subject to the relation ; and the envelope of this line , or " " alien\ldquo ; chord , may be found for any assigned form of In the case when is J. J. ON THE DIAMhTERS OF A PLANE CUBIC .
and since The equation of the " " ] \ldquo ; chord is then simply Vith ( in virtue of which the equation of the " " alien\ldquo ; chord reduces to ; . . .
( 87 ) the envelope of which is , . . .
( 88 ) a conic touching the line at the point which is harmonic-conjugate to ( x'y'O ) with respect to the intersections of with the poloid ; and touching the polar line of at the point in which it is met by , the point : : : : ' \mdash ; by'3 , , at the -point on that polar line ( 79 ) , S48 .
62 .
The conic envelopejust found may be called the satellite-conic\ldquo ; to the polar line of x'y'z ' .
Of course , this conic touches also the double chord of the pencil through x'y'O , since it is at once all " " alien\ldquo ; and a proper chord of the polar line of ( x'y'O ) .
Arr nging the equation ( 88 ) to the satellite conic as a quadratic in it is giving at once , for the envelope of the system of conics satellite to the complex of polar lines of the points on , the quartic which developed is the envelope of the satellite conics is the system of four lines and the tangents to at the points , S49 .
The explanation of this is that at the point of intersection of the polar line ( 76 ) MR. J. J. WALKER ON THE DIAMETERS OF A PLANE CUBIC .
with one of the tangents ( 66 ) , that is ] the " " alien\ldquo ; chord for that point : for ( 76 ) ( G6 ) give : : , and the substitution of these values in ( 87 ) gives for tlJe " " alien\ldquo ; chord , after dividing out the extraneous factors , simply ; , the tangent to at the point : Thus the system of satellite conics is inscribed in the quadrilateral formed by the line and the three tangents to at the points in which meets it .
A satellite bola is shown in Plate 8 , when is the line at infinity .
63 .
The tangents to from the pole of the transversal being real only when meets the cubic in a single real point , it is desirable to use as lines of reference another pair in connexion with , which shall be real in all cases for purposes in which their reality is essential .
Consider the connector of the pole of with a real point ( A ) in which meets , taking it , say , as ; and let the -polar of the point , be taken as the third line ( BC ) of reference , .
The equation of is then reduced to the three terms , remembering that still , . . . .
( 90 ) since the triangle is self-colljugate with pect to that conic ; and as conditions for the terms disappearing from its equation ( 91 ) But since the equation of the three tangents to at the points is now of the form and one of them passes through the point the term must disappear fro ] ] ] , or which ( 91 ) necessitates either or ; but the supposition would impl .
a second of the tangents to at the points passing through the point , so that therefore , by the second of the three conditions ( 91 ) ; whence , by the first , .
( 92 ) 64 .
Thus the equation to the cubic is reduced to the terms , ( 93 ) with the relation ( 92 ) ; and ( 90 ) the poloid to or ( 94 ) where \ldquo ; ( 95 ) or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or where \ldquo ; or the reciprocal of which is C5 .
For the above forms ( 93 ) ( 94 ) of ?
and , the values of their first differential ients , with the coordinates of a point on substituted for , are ( 96 ) MR. J. J. ON THE DIAMETERS OF A CUBIC .
giving ( 36 ) for the -point on the -polar line of that point : : : : \mdash ; . .
( 97 ) The values give ( 9S ) in virtue of the relation ( 92 ) ; , the above is the equation of the Cotesian for the form ( 93 ) of now used .
66 .
The Cotesian for the forms of at present employed has been found quite independently of any application of the general equation ( 37 ) given above ; but it may be of interest to test the general formula by this result .
The only terms in the general expression which do not disappear in virtue of are four times which , taking the values of , for the present case , is identically equal to four times .
Rejecting the factor , and applying the relation ( 92 ) , , or this result is , as before ( 98 ) ; or , with the rejected factors restored , , ( 98 bis ) a cubic having a node at the point , the -pole of or ; the nodal tangents being or ( 92 ) ( 99 ) The node , therefore , will be or ?
, as are of the same , or unlike signs .
67 .
Since , now ( 94 ) it appears that the nodal tangents ( 99 ) of are the tangents to at the points , or , and will therefore only be real when cuts its poloid in real points , as shown before .
The transversal meets the Cotesian , as well as the cubic , at the points , . . .
( 100 ) which last two will be real only when is , or when does not meet in real points .
68 .
The tangents to at the points ?
, or , are ( 48 ) if , above , ; viz. , in this case , then , , ( 101 ) since , for the above form of , with . . . .
( 102 ) Now , the within brackets in are equal to or ( 103 ) MR. J. J. WALKER ON THE DIAMETERS OF A PLANE CUBIC .
which factors may be identified with the three tangents to at the points ; viz. , the points ; These will be real points only when .
are of unlike sign ; , when does not meet in real points , and is acnodal .
69 .
The Cotesian ( 98)\mdash ; discarding the factor and the primitive cubic are plainly intersected by , or , in the same three points , and the tangents to ) at these points will be derivable from those to by changing therein into into ; viz. , their equations are . . . . .
( 104 ) 70 .
The equation to being shows by its form that the tangent to at the point ( or ) viz. , the line ( 103 ) or touches at the point 71 .
Again , throwing the equation to into the form the terms within brackets being ( 103 ) the tangents to at the other two points at which meets it , it appears that these lines touch on the line i.e. , at the points determined by it , and , . . . . . . .
( 106 ) the polars of those two points of the three common to and For these points are determined ( 9S ) by and the polars of which , with respect to ( 94 ) are or 72 .
The equation of the double chord ( S41 ) of the pencil through x\fnof ; m on being generally ( 56 ) is , in the present case , from the values of the differential coefficients given ( 98 ) , S65 , , . . .
( 107 ) or , arranged as a binary cubic in The discriminant of this last gives the envelope of the double chord as the point describes the line , or ; viz , , it is ( 108 ) or in virtue of the relation ( 92 ) MR. J. J. WALKER ON THE DIAMETERS OF A PLANE CUBIC .
The former value of shows that the polar of , one of the three points common to and , viz. , is the tangent at a cusp at the point , . . . . .
( 110 ) the tangent meeting ) for the fourth time at the point , or . . . . .
( 111 ) 73 .
Now , since any real intersection of and ?
might be taken as the point , it follows that when these three intersections are real there will be three real cusps to , the cuspidal tangents being the polars , with respect to the poloid , of the points common to and This is shown , independently , for the two intersections of and other than , by the second form of , which gives simultaneously ( 109 ) ( 112 ) of which the former has been shown to be the ( square of the ) -polars of the two intersections in question ( 106 ) .
74 .
It is plain from the forms of that the triad of coordinates ( 110 ) , ( 112 ) satisfy the first differential coefficients of ; in fact these are , , Also , giving for the cuspidal tangent when in .
are introduced the values as in ( 110 ) ; when the values as in above .
The same values substituted in the second differential coefficients will be found to make 75 .
Considering the intersections of and , the elimination of .
between their equations leads to or . . . . . .
Referring to the equation of , S72 , it appears at once that or ( 92 ) gives , the tangent to at the point , ( 103 ) , S68 , touches at the point in which meets ; and the same line has been shown ( S70 ) to touch at that point .
From this it would follow at once that the other two tangents to , at the points or , touch both and at the same points ; but this is shown by ( 113 ) above , since and appear at once from it to have double contact at the points in which is met by the line MR. J. J. WALKER ON THE DIAMETERS OF A PLANE CUBIC .
which has been proved ( S 71 ) to be the chord of contact with of the two tangents referred to 76 .
Generally then , when the transversal meets the cubic in three real points , the envelope of the double chords of pencils of lines through points in , is a tricuspidal quartic having triple contact with , the poloid of , at the points in which it is touched by the three tangents to at the points ; the cuspidal tangents being the -polars of those three points ; and the fourth pair of points common to and lie on the line L. In Plate 8 the envelope is figured for the case of being the line at infinity and the cubic a modification of the Cissoid , as more particularly described below ( S91 ) .
The property of the double chord touching in a point which is harmonic to its second intersection with the poloid relatively to its -point and intersection with , has been proved in S59 by the use of other lines of reference , with the reality or imaginariness of which it is , plainly , unconnected .
77 .
When touches , it and the tangent at the point where it again meets the cubic being taken as and their chord of contact with the poloid as the equation of the poloid ( 10 ) is reduced , since , to the conditions for which are , . . . . .
, . . . . . . .
( ii .
) , . . . . . . .
ab \mdash ; . . . . . .
But since is on ; therefore , by ( iv .
) , since is excluded by the form of ; therefore , by , and by ( iii .
) finally , by ( ii .
) ; since is excluded by the form of Thus , it appears that the cubic and poloid are reduced to , ( 114 ) . . . . .
( 115 ) The form of shows that passes through the point in which touches it ; or the poloid touches the cubic at the point of contact of the transversal L. 78 .
The coefficients of the reciprocal of which do not vanish are ' ; which values , with , give ( 37 ) , S80 ( dividing out ) , ; . . . . . . . . .
( 116 ) the locus degenerating in this case into the line and a conic having double contact with .
at the points where the -polar of the point of intersection of and meets it and 79 .
The double chord through any point on or , in this case , wherein is or , ( 117 ) MR. J. J. WALKER ON DIAMETERS OF A CUBIC .
the envelope of which as describes the line or , is ( 118 ) another conic having double contact with and at the same points .
80 .
Lastly , the chord rough a point on the polar line of , any point on which has the former as its CoTEs-point , but which is not one of the pencil of chords through ( x'y'z ' ) , is ( 53)\mdash ; when , and is But here ; whence the chord in question is . . .
( 119 ) 81 .
The polar line of x'y'O being the envelope of the chord ( 119 ) above , as describes the polar line just mentioned , determined as the condition that shall touch is\mdash ; dividing out the factor , . . . . .
( 120 ) or , as it may be otherwise written , , ( 121 ) a conic inscribed in the triangle formed by the or , and the polar line of the point on L. WTONIAN DIAMETERS .
82 .
If is the line at infinity , then : : : : : : ; and to the -point of a chord through a givenl point on ( x'y'z ' ) corresponds the " " mere ] point\ldquo ; on a chord parallel to a given line , the " " direction coordinates\ldquo ; of which are , S 44 , ; . . . . .
( 122 ) the line being written ; viz. , the quantities ( 122 ) may be considered as the coordinates of the point in which this line , and all parallel to it , meet the line at infinity .
To every equation in the preceding part of this Paper involving as parameters ( x'y'z ' ) the coordinates of a point on , a finite line , now corresponds one in which x'y'z ' are replaced by : : : . . . .
( 123 ) Thus , for the polar line of now appears , S 17 , i. e. , , ( 124 ) which is , plainly , the locus of a point on a chord parallel to , and meeting the cubic in the points , such that ; . . . . .
( 125 ) viz. , is the .
' mean point\ldquo ; on the chord relatively to the triad The line ( 124 ) is the Newtonian Diameter of the system of chords ; and its envelope , since ( 12.3 ) . . . . . . . . .
( 126 ) the " " centroid i.e. , the ' poloid\ldquo ; of the line at infinity , is . .
( 127 ) or where is now the condition should be a parabola . .
, MR. J. J. WALKER ON DIAMETERS OF A PLANE CUBIC .
, that the line at infinity should be cut harmonically by the conics The condition that the cubic should touch the line at infinity in the standard form is equal to four times that for the centroid being paraboJic and touching the finite asymptote of ; and 83 .
When the cubic meets the line at infinity in three real points , or three asymptotes are real , the centroid is an ellipse inscribed in ) the trlangle formed by these lines , so as to touch them at their middle points ( Plate 8 ) .
But if the cubic has ] one real intersection with the line at infinity , then the centroid is a hyperbola , having the single real asymptote as a tangent ; and The asymptotes of the hyperbolic centroid are the only real pair of conjugate meters of the , viz. , each cuts every chord parallel to the other in its mean point , and in particular divides the tangential chord of the cubic parallel to the other , or the parallel nodal chord if the cubic is nodal , in the ratio of 2 : 1 .
Thus in Plate 7 the chords parallel to one asymptote of the centroid , are divided in such wise by the other asymptote in the points 84 .
The mean point on any diameter of regarded as a chord of that cubic is the point in which it is met by the diameter of the centroid conjugate in direction to its chords , the " " double\ldquo ; one ( S 7 ) of which is the -polar S 42 of that mean point .
The locus of the mean points of diameters of the cubic is the noda ] cubic , the Cotesian of the line at infinity , having as its nodal tangents the asymptotes of the centroid , and being therefore acnoded when the three asymptotes of the cubic are all real ( Plate 8 ) , the acuode , or conjugate point , being the centre of the elliptic centroid .
The asymptotes of the cubic meeting , two and two , on the diameters of through its points of triple contact with , are parallel to those of the primitive cubic ; and the line at infinity is its inflexional axis , these inflexions being at the points in which the asymptotes of meet that line .
In Plate 8 the Cotesian is represented with three hyperbolic branches , each touching the centroid ; in Plate 7 , with a real loop touching the centroid , and a single real asymptote parallel to that of the 85 .
The envelope of the " " double ordinates\ldquo ; of Newtonian Diameters is represented in Plate 8 as a quartic ( w ) with three real cusps , the three diamekrs of the centroid conjugate to the asymptotes of being the cuspidal tangents .
This quarti as well as the Cotesian , has triple contact with the centroid at the points of contact of the asymptotes of .
In Plate 7 , the cubic having only one real asymptote , the quartic has only one real cusp ; also it and have only one contact with at the point of contact of the real asymptote with .
Within the limits of the diagram the only parts of visible are that adjoining the contact referred to , terminating on one side with the point of contact ( C ) of the double ordinate CD of the diametel \mdash ; in this case ( of being at infinity ) the point on CD conjugate to ( S59 ) being at infinity , is the centre of the segment CC'\mdash ; and ( ii .
) the cusp of with the contact of another double ordinate , of the diameter being the centre of .
It be observed that are the mean points of the diameters being harmonic conjugate to with respect to , the other two real intersections with .
86 .
In Plate ( 8 ) is shown an " " alien\ldquo ; ordinate of , having its meaI ) point on that diameter ( S 61 ) , but being an ordinate of the other tangent to which might be drawn through .
The parabola , which is the envelope of these " " alien\ldquo ; ordinates through ths different points on , is also shown , touching at the point , and having the connector of that point with the centre of the centroid as diameter ( 88 ) , this connector being now the representative of the of the equation just cited , the connector of the -point on the polar line of ( x'y'O ) with the pole of the transversal , now become the line at infinity .
87 .
The asymptotes of the centroid offer themselves as an unique pair of Cotesian axes to which the cubic may be frequently referred with advantage , its equation being then , . . . . .
and the centroid , with changed sign\mdash ; which is now immaterial\mdash ; . . . . . .
( u. ) For lnany discussions also it is convenient to define the diameter by the coordinates of its ) oint of contact with The equation of its double ordinate will now be ; . . . . .
( iii .
) and that of its " " alien\ldquo ; ordinate at the point . . . . . .
It will be sufficient to indicate steps by which these may be verified independently of the general formulae given in the earlier part of this Memoir .
The diameter is the tangent to ( ii .
) at or , . . . . .
the coordinates of the mean point of which may be found as one-third of the sums of those of its intersections with .
The polar of this mean point is ( iii .
) ; and that of the mean point of the second tangent to through on is ( iv .
) MR. J. J. WALKER ON DIAMETERS OF A PLANE CUBIC .
88 .
The secondary chords through points lying on a given diameter , and having them as their mean points , being inclined at varying angles with that diameter , to determine that one which makes a given \mdash ; with the diameter being the angle between the asymptotes of the , it is the chord which meets the diameter in the point such that : : In particular , the point at which the secondary chord is perpendicular to the diameter is determined by .
The angle between a diameter and its primary chords or ordinates is given by and those diameters which are perpendicular*to their chords are the angents to the centroid at the points determined by i. e. , the intersections of the conic with the centroid ; or , multiplying by , those of the The finite intersections in question are only three in number , their coordinates being determined by the cubics 89 .
When the asymptotes of the centroid are imaginary , two convenient axes of coordinates are found in the diameter parallel to one asymptote of and its * Diameter autem ad Ordinatas rectangula si modo aliqua sit , etiam Axis dici potcst NEWTON , ' numeratio Lincarum Tertii Ordinis , ' S 2 .
; to which lines the other two asymptotes will form a conjugate harmonic axes are the special form which those of the self-conjugate triangle of reference employed , S63-76 , take ; and all the equations there deduced will be adapted tu bove Cotesian axes simply by substituting unity for 90 .
When the cubic is parabolic , i.e. , touches the line at finity , its " " centroid\ldquo ; also touches that line , viz. , it becomes a parabola , which may then be referred to the tangent parallel to the finite asymptote of , and the diameter through its point of contact .
The Cotesian ?
2 will then be made up of the line at infinity and a parabola , and the ords of having their mean points on a diameter of which they are not ordinates will another bola , while the double chord will envelope a third .
In this case all the equations formulae of S77-81 are applicable by simply making It has not been necessaly to add a figure in illustration of this case , the curves all of character .
0F PLATES .
91 .
The two types of cubic which have been drawn to illustrate results arrived at in this Memoir have been constructed geometrically with great accuracy as follows:\mdash ; Those in Plate 6 and Plate 8 from two diameters of a hyperbola , as the well-known Cissoid of DIOCLES from two rectangular diameters of a circle ; from a vertcx of one diameter a pencil of lines was drawn , each to the extremity of an ordinate parallel to the other diameter , and then its intersection with the equidistant ordinate on the other side of that diameter determined a point the locus of which the cubic represented , with one cuspidal branch and two hyperbolic .
For the in the line was drawn arbitrarily , cutting in three real points , A and the ngents to at these points traced , the accuracy of their directions by their tangential points proving to range in a right line poloid was next inscribed in the triangle formed by the triad of tangents , oints A , harmonic-conjugate severally to with respect to the lers of the triangle .
Another al.bitrary line OD having been drawn touching the conic in and .
the cubic in three real points , its -point was found by actual calculation from the measured lengths of the segments between and the cubic 21 .
The polar of , being the double chord of the pencil through the point on of which OD was the polar line , of course determined that point .
[ Otherwise crht have been taken arbitrarily and the -point on relatively to it ] been found by measurement ; and so the polar-line of with the double chord ] ) been arrived at .
] The accuracy of the figure so far was tested by ex the reement of the -points of itself and chord with their ersections by the polar line at For the figure in Plate 8 , corre- .
J. J. ON THE ) A PLANE CUBTC .
20.3 " " \ldquo ; ibed in the by the of so as to bisect the sides .
) Cotesian and the quartic were down from their equations without difficulty , ) triple contacts with , and the ] of the mptotes of to those of , combined with ] the simply defined positions of the cusps , and directions of the cuspidal tangents of ) , enabling them to be traced with great accuracy .
The parabolic envelope shown in connexion with the diameter touching that line at its point having an axis parallel to the connector of that point with the centre of the centroid , as well as touching the asylnptotes , readily constructed geometrically from data , nd its ) tested by tangent drawn arbitra.rily , meeting in three real points and cut i ) in to have the last as its mean point , as by actual measurement of segments , with great exactness .
( ii .
) For the second illustration , Plate 7 , the figure of a cubic only a real asymptote was obtained by constructing a Cissoid two diameters of an ellipse , precisely as that of DIOCLES from the , of which in fact well its the figure previously described , may be arded as projections .
The hyperbolic centroid was then constructed ) its equation referred to the two conjugate diameters of the generating ellipse , which are also conjugate of the centroid , and to one of which the real asymptote of is parallel .
The Cotesian with its real single real asymptote , parallel to that of also the asymptotes of the centroid as nodal tangents ( S 50 ) , was constructed from its equation , the loop toucbing the ce1ltroid at the contact with it of the real asymptote of .
It will } ) remarked in this as in the case preceding where they are all three real , how soon the cubic approaches its asymptote , and its becolnes .
As the little of visible within the limits of the figure ) remarks have already been made , S 60 .
Two diameters , besides the conjugate pair which the asymptotes of form , S 83 , have been introduced , ouching one branch of and three intersections with ) touching the other branch of and in only one , its mean point , .
The " " double\ldquo ; chords of these of .
mean points , toucbing their envelope ) at the points , equidistant from respectively as , are typical of kind .
It will be that they are " " double\ldquo ; chords of , with which , however , in this , they ] only one intersection .
The mean point of the chord , drawn parallel to the donl , chord , as determined by actual nent of ments , ides with , its intersection with the diameter , to a nicety .
Sotne remarks have already been made , , on the ] parallel to one asymptote of and cub by the other in their mean points with great exactness .
It not been desirable to introduce the r.abolic envelope connected with either of the meters drawn , into figure .

