Full text: The collected mathematical papers of Arthur Cayley, Sc.D., F.R.S., late sadlerian professor of pure mathematics in the University of Cambridge (Vol. 8)

392 
ON THE TRANSFORMATION OF UNICURSAL SURFACES. 
[523 
The necessity for the term co appears by the consideration that if we apply to 
the plane figure a Cremona-transformation, thus obtaining a new transformation of the 
surface, the value of 2a,. will in general be altered ; whereas the expressions for q, t 
should it is clear remain unaltered; and it arises as follows, viz. for certain transfor 
mations of the surface the curve of the order (k — n 4- 4) n — 3, passing (k — n -t- 4) r — 1 
times through each point a r and assumed to be the projection of the residue, is not 
an indecomposable curve but contains a certain number co of factors (each belonging 
to a unicursal curve definable by means of the number of its passages through the 
several points a r ), which factors are to be rejected in order to obtain the equation of 
the proper residue. Thus reverting to the transformation 
x : y : z : w = x'* : y'* : : ix' + y' + /) 2 
of Steiner’s surface, the projection of the quadric residue was (as already remarked) a 
line; applying to the plane figure the ordinary quadric (or inverse) transformation we 
introduce three fixed points, (a 2 =3), say these are A, B, G\ viz. in the new trans 
formation of the surface the projection of any plane section is a quartic curve having 
a node at each of the fixed points: the projection of the residue ought clearly to be 
a conic through the three points; but according to the general formula it is a quintic 
having at each of these points a triple point: the quintic is in fact made up of the 
lines BG, GA, AB and of the conic which is the proper residue; viz. in the case in 
question there are 3 factors thrown out, or we have co = 3. To apply this to the 
second investigation of 2q + 91, by comparison of the two deficiencies, observe that in 
general if a curve is made up of co +1 indecomposable curves, the deficiency of the 
compound curve is equal to the sum of the deficiencies of the component curves — co; 
hence if co of the curves are unicursal, the deficiency of the compound curve is equal 
to that of the remaining curve — co ; or, what is the same thing, the deficiency of the 
remaining curve is = that of the compound curve + co; and the addition of the term 
+ to to the expression for the deficiency is thus accounted for. It is easy to see that 
a like explanation applies to the first investigation of 2q + 91. 
I further remark, reverting to the equations 
x : y : z : w = X' : Y' : Z' : W' 
of the transformation, that the product of the co factors is given as the common factor 
(if any) of the Jacobians 
J(Y\ Z', W'), J(Z', W', X'), J{W, X\ Y') and J(X', Y', Z'). 
Such common factor exists whenever we can by a Cremona-transformation of the plane 
figure reduce the number of the points a r upon which the transformation of the 
surface depends; viz. for any given transformation of the surface, co is equal to the 
excess of above the minimum value of Xa r , or, what is the same thing, 2a,. — co 
is equal to the minimum value of 2a r , and is thus independent of the particular trans 
formation. And of course if 2a r has this minimum value, viz. if the transformation 
is such that the number of the points a r cannot be reduced by any Cremona-trans-
	        
Waiting...

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.