I found this in the first comment: "Note that only *electro***static** field lines cannot form closed loops. If the magnetic flux is changing, it will induce a current via closed loops of electric field. – " at https://physics.stackexchange.com/questions/105592/why-cant-electrostatic-field-lines-form-closed-loops

The motivation for it is that I noticed in a square grid like graph paper, and on a cellular automata with a square grid (an explosive rule in a normal automata like Life, but explosive so that the cells flicker on and off across the entire screen and fill the screen), is that if you measure a cell with a ruler, the diagonal length is root 2 times longer than the edge, but the counting of the diagonal cells in a group of 9 in a square grouping, is the same as the two legs (or the counting length of cells in a 45-45-90 "triangle" is the same as the legs). However, when zoomed out in golly, if the eye sees something moving, diagonal elements travel faster than vertical or horizontal motion, due to a constant update with no preference in terms of the speed that a cell turns on or off, so that a diagonal moving group of cells must travel faster because they have to travel a longer distance in terms of real distance. I thought that by having a variable increase in how fast a cell turns on, ie it turns on faster if a cell is born to the right of a cell. Then (as it stands with no variable it causes small rotations in small collective cells), with this way, since cells are born faster and propagate faster towards the right horizontal direction, there is now a bias in a vector to the right and a positive or negative vector along either diagonal. If you draw a vector sum of a vector to the right and a diagonal vector, before the diagonal vector was longer since you before didn't have bias in the right direction as well, so that now, the "bias" is now only in the up-down direction. So there should be a separation of elements in the up-down direction as opposed to the diagonal, so instead of circular rotations, there might be planar twisting of rotations. It may be both up-down and left right ie a cell directly above (not one above and to the left or right a diagonal cell) cells turn on faster. Also, to generate the correct diagonal (1.4 ... an irrational number) the faster update (the range of cells), may need to be continuous or linked to an analog crystal in an arduino chip like this: https://www.fpga4fun.com/oscillators.html connected to the computer, to generate the correct irrational update (as well as handling the range of updates, and how to make the update rules for the automata).