00:01
That the dihedral group of order 8 has two isomorphic subgroups of order 4 and show that there exists a 2 to 1 mapping homomorphism from the quaterning group to d4 onto these either of these subgroups so really either of the subgroups or both of the subgroups rather i should say and then find what the kernel is of that mapping so first let's look at the multiplication table 4d4 this is actually missing one of the subgroups here so here this this yellow subgroup is going to be one of the ones of interest and the other one which kind of will make sense is going to be this one sorry this this one here this subgroup so we'll have the yellow subgroup and we'll have this purple subgroup where these two subgroups represent flips about two orthogonal axes as well as a rotation by pi over two of the square, which we could also consider as a rotation about another orthogonal axis.
01:24
So here we'll have like the two principal axes, and we would think of the other ones as the diagonal axes, and then it becomes pretty clear to see how those operations are going to be isomorphic.
01:42
How do we show that they're isomorphic? well, we could draw a picture, or we could just look at the separate multiplication tables for those two subgroups of d4.
01:54
So if you look at these, so here i have f1 and f2 as the main axes of the square, and then f3 and f4 are flips about the diagonal axes, and then q is that rotation by pi of the square in the plane.
02:20
So just looking at the tables, you can see that there is, in fact, an isomorphism here.
02:26
But now the question remains, how do we construct this 2 to 1 mapping from the quaterning group? so if we look at that multiplication table, we can kind of see that if we group the positive and negatives together, we get this same structure.
02:46
We get the same like reverse diagonal of that axis, and then considering negative 101 is the same.
02:57
Identities across the main diagonal.
03:00
So we have that structure.
03:02
So we can see visibly how it's there, but how do these things actually relate? so the homomorphism that i would suggest, the mapping that makes the most sense, would be to use rotations.
03:19
So if we consider each of these as rotations, so even the flips about the axes as rotations, so here are these, or rotations by pi about let's say the q axis.
03:40
Where really here q, we would want to label these.
03:45
So let's label these as, let's label this.
03:48
Really here, this is more like the z axis in normal pictures, but here let's consider this as the i axis.
03:55
And then if we label this as the j, the j axis.
03:59
And this as the k -axis.
04:06
We can see, so that would require us then to represent each of the quaternian elements...