3 ms·
No, imagine a licorine candy https://www.google.com/search?q=licorice+candy&tbm=isch https://www.google.com/search?q=licorice+candy&tbm=isch but the cross secti
by gus_massa 6mo ago
No, imagine a licorine candy https://www.google.com/search?q=licorice+candy&tbm=isch https://www.google.com/search?q=licorice+candy&tbm=isch but the cross section is a +.
(It's easier if you get one that is not twisted. Something like https://www.sugarrushlane.com/products/filled-red-licorice-swedish-candy?srsltid=AfmBOopOhrX8c1uvrtQ4PD_ChEebZGHdJ-YMP2QemgHu4KwvcPVxXN9l https://www.sugarrushlane.com/products/filled-red-licorice-s... but with 4 instead of 5 parts. Also imagine each of the 4 parts has a different color. )
Now make a circle with it.
If you don't twist it, you get an horizontal surface that looks like a thick border of a circle, and a vertical surface that looks like the round wall of a cylinder.
If you twist half a turn, when you connect the extremes the top of the + in one extreme is glued to the bottom of the + in the other extreme. Something similar withe the left and right parts of the +. Now you have two overlapped copies of a Möbious strip, each perpendicular to the other.
If you twist a quarter of a turn, when you connect the extremes the top of the + goes to the right, the right to the bottom, the bottom to the left an the left to the top. So you get only one "strip?"
- lupire 6mo agoNote that the colors are important. You need a discrete structure with N rotational segments and a turn of 1/M. Otherwise any rational turn size will have have multiple Mobius strips.