Families of cyclic cubic fields... and more

carpetpool · 2018-07-25, 05:12

It is well known that if K is the splitting field of the polynomial t = x^3 + y*x^2 + (y-3)*x - 1, then K must be a cyclic cubic field, as explained here. This result is known as the "simplest cubic fields". However, I was interested in finding more cyclic cubic families, and I did happen to spot another occurence:

If K is the splitting field of the polynomial T = x^3 - y*x^2 - 9*x + y (assume t is irreducible), then K must be a cyclic cubic field. The proof for this is similar to the one illustrated in the pdf link. The discriminant of t is d = (y^2 - 3*y + 9)^2 while the discriminant of T is D = (y^2 + 27)^2. Both splitting fields of y^2 - 3*y + 9 and y^2 + 27 are isomorphic to the field of Einstein Integers --- a field obtained by adjoining a root of y^2 + y + 1 to the field of real numbers, known as the third cyclotomic field.

Besides these two sole examples, are there any more unique cyclic cubic field "families" or is there perhaps a finite number of instances, and it might be the case that these are the only two instances?

For quartic cyclic field families,

by trial and error, I found that when K is the splitting field of the polynomial T = x^4 + y*x^3 - 6*x^2 - y*x + 1, then K must be a cyclic quartic field. t has the discriminant D = (y^2 + 16)^3.

The same question related to cubic fields is addressed to quartic fields.

carpetpool · 2020-04-11, 02:19

I've been back on researching the task again. I thought I might share some new results, if they are not known already:

x^5 + (n^2 + n + 4)*x^4 + (-2*n^2 - 2*n + 2)*x^3 + (-n^4 - 2*n^3 - 10*n^2 - 9*n - 5)*x^2 + (-3*n^2 - 3*n - 2)*x + n^2 + n + 1

x^7 + (n^3 - 3*n^2 - 4*n - 1)*x^6 + (-3*n^5 + 9*n^3 + 3*n^2 + 3*n - 12)*x^5 + (3*n^7 + 6*n^6 - 4*n^5 + n^4 - 7*n^3 + 28*n^2 + 15*n + 7)*x^4 + (-n^9 - 5*n^8 - 2*n^7 - 5*n^6 + 6*n^5 - 11*n^4 - 21*n^3 + 12*n^2 - n + 28)*x^3 + (n^10 + n^9 + 2*n^8 - n^7 - 9*n^6 + 4*n^5 - 32*n^4 - 6*n^3 - 49*n^2 + 5*n - 14)*x^2 + (-n^9 + 3*n^8 - n^7 + 11*n^6 + 4*n^5 + 4*n^4 - 10*n^3 - 30*n^2 - 20*n - 9)*x - 3*n^7 + 3*n^6 - 7*n^5 + 14*n^4 - 5*n^3 + 12*n^2 - 6*n - 1

As far as I know, there is only one other known cyclic septic besides the one I have listed here.

Dr Sardonicus · 2020-04-12, 01:04

I believe your cubic t should be x^3 + y*x^2 + (y-3)*x - 1 (not +1). Then, replacing y with -y gives the usual formulation for the simplest cubic fields.

The discriminant of your cubic T is 4*(y^2 + 27)^2.

If you replace y with -y in your quartic T, it will coincide with the usual form for the "simplest quartic fields."

You might want to look up Washington's cyclic quartic fields.

I haven't tried comparing with yours, but there is a family of cyclic degree-5 polynomials known as Emma Lehmer's quintics.

There is a family of "simplest" degree-6 number fields, defined by a one-parameter family of polynomials of degree 6 which have cyclic Galois groups.

2018-07-25, 05:12	#1
carpetpool "Sam" Nov 2016 5·67 Posts	Families of cyclic cubic fields... and more It is well known that if K is the splitting field of the polynomial t = x^3 + yx^2 + (y-3)x - 1, then K must be a cyclic cubic field, as explained here. This result is known as the "simplest cubic fields". However, I was interested in finding more cyclic cubic families, and I did happen to spot another occurence: If K is the splitting field of the polynomial T = x^3 - yx^2 - 9x + y (assume t is irreducible), then K must be a cyclic cubic field. The proof for this is similar to the one illustrated in the pdf link. The discriminant of t is d = (y^2 - 3y + 9)^2 while the discriminant of T is D = (y^2 + 27)^2. Both splitting fields of y^2 - 3y + 9 and y^2 + 27 are isomorphic to the field of Einstein Integers --- a field obtained by adjoining a root of y^2 + y + 1 to the field of real numbers, known as the third cyclotomic field. Besides these two sole examples, are there any more unique cyclic cubic field "families" or is there perhaps a finite number of instances, and it might be the case that these are the only two instances? For quartic cyclic field families, by trial and error, I found that when K is the splitting field of the polynomial T = x^4 + yx^3 - 6x^2 - yx + 1, then K must be a cyclic quartic field. t has the discriminant D = (y^2 + 16)^3. The same question related to cubic fields is addressed to quartic fields. Last fiddled with by carpetpool on 2020-04-14 at 21:56*

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Lucas number cubic and quadratic reciprocity	carpetpool	Miscellaneous Math	0	2018-06-27 03:12
Cyclic fields with class number h	carpetpool	Abstract Algebra & Algebraic Number Theory	0	2018-01-30 06:10
Efficient computation of cubic residue	axn	Computer Science & Computational Number Theory	2	2015-01-04 03:49
visualizing prime families	Wini	Aliquot Sequences	5	2011-04-22 20:05
Intersection with cubic	kuratkull	Miscellaneous Math	7	2008-01-24 15:28

2020-04-11, 02:19	#2
carpetpool "Sam" Nov 2016 14F₁₆ Posts	I've been back on researching the task again. I thought I might share some new results, if they are not known already: x^5 + (n^2 + n + 4)x^4 + (-2n^2 - 2n + 2)x^3 + (-n^4 - 2n^3 - 10n^2 - 9n - 5)x^2 + (-3n^2 - 3n - 2)x + n^2 + n + 1 x^7 + (n^3 - 3n^2 - 4n - 1)x^6 + (-3n^5 + 9n^3 + 3n^2 + 3n - 12)x^5 + (3n^7 + 6n^6 - 4n^5 + n^4 - 7n^3 + 28n^2 + 15n + 7)x^4 + (-n^9 - 5n^8 - 2n^7 - 5n^6 + 6n^5 - 11n^4 - 21n^3 + 12n^2 - n + 28)x^3 + (n^10 + n^9 + 2n^8 - n^7 - 9n^6 + 4n^5 - 32n^4 - 6n^3 - 49n^2 + 5n - 14)x^2 + (-n^9 + 3n^8 - n^7 + 11n^6 + 4n^5 + 4n^4 - 10n^3 - 30n^2 - 20n - 9)x - 3n^7 + 3n^6 - 7n^5 + 14n^4 - 5n^3 + 12n^2 - 6*n - 1 As far as I know, there is only one other known cyclic septic besides the one I have listed here.

2020-04-12, 01:04	#3
Dr Sardonicus Feb 2017 Nowhere 2·3²·5·71 Posts	I believe your cubic t should be x^3 + yx^2 + (y-3)x - 1 (not +1). Then, replacing y with -y gives the usual formulation for the simplest cubic fields. The discriminant of your cubic T is 4*(y^2 + 27)^2. If you replace y with -y in your quartic T, it will coincide with the usual form for the "simplest quartic fields." You might want to look up Washington's cyclic quartic fields. I haven't tried comparing with yours, but there is a family of cyclic degree-5 polynomials known as Emma Lehmer's quintics. There is a family of "simplest" degree-6 number fields, defined by a one-parameter family of polynomials of degree 6 which have cyclic Galois groups.