Ray Keller at UVA has done the best research on this.;
Somehow, the individual cells get flattened, like a stack of coins.

Collagen fibers get wrapped around the cells in many pairs of helical spirals (each with fibers oriented 90 degrees to the others)

The physical strength of the notochord in tadpoles & fish results from counter-balanced forces of tension in these collagen fibers versus swelling of vacuoles in the cells.

Personally, I tend to think that the alignment of this collagen must also be part of what molds the shape of the notochord; but I can't invent a good experiment to test this.

Some medical textbooks, claim that the notochord sheath becomes the sheath of the intervertebral discs. Probably they are mistaken, and the notochord degenerates, and the intervertebral discs develop where it was.

"Slipped" disks are really tears (herniations) in the disk sheath. This sheath is made of alternating perpendicular layers of collagen that are oriented in the same spiral direction as in the notochord.

Thestrips of mesoderm on either side of the notochord are called the paraxial mesoderm. Most of it segments to form somites.

In the head, this paraxial mesoderm does not segment:
This head paraxial mesoderm is called the prechordal plate.

Except supposedly shark embryos form a few head somites?

Muscles, connective tissue, and some head bones form by differentiation of prechordal plate cells.
For example, the muscles that rotate the eyeball. Some head skeleton is formed, instead, by neural crest.

Scanning electron microscopy shows that these cells rearrange to form swirls, (named "somitomeres") just before they actually segment into somites, which begin as hollow epithelial balls.

Becoming epithelial may be part of what causes somites to split apart?
Or maybe it is some loss of adhesion?
Or maybe there are circumferential constrictions?

The physical cause of somite separation is NOT yet known.

Somites only last for part of embryonic development; their cells soon rearrange and disperse.
But the locations of the somites controls where later structures will form, for example the bones of the back-bone (vertebrae).

This was proven by grafting 5 (small) somites from the tail to replace 3 of the (larger) somites of the trunk (using salamander embryos in the 1920s).

In frog embryos, supposedly the somites are never epithelial, & their segmentation is (?) by 90 degree rotations of aligned cells
At first these are lined up middle-to-side and then rotate to anterior-posterior direction in separate blocks.

The embryologists who discovered this difference were very good!

(It could be that what they saw was myotome formation?)
How plausible is it that completely different physical mechanisms are used to make somites in frogs, as compared even with salamanders?

The number of vertebrae and spinal nerves in the body are controlled by the number of somites formed in the embryo.

In humans, this number is tightly controlled, (31, 32, or 33)

In some species of fish, this number varies with temperature.

But however many pairs of somites are formed in an embryo, then that's how many bones it will have in its backbone.

Experiments have been reported in chickens in which cutting off the hind embryo caused formation of smaller somites. so that the total number of somite pairs turned out the same.

But I have also read reports that say we form more somites that we need, and then any beyond 31-33 degenerate! Which is the truth, I would like to find out.

Most current embryologists give no thought to physical causes or mechanical forces. Our textbook regards the "noggin" protein as being the cause of somite formation, because...

* fluorescent antibodies show that noggin is concentrated there

* putting extra noggin protein on lateral mesoderm causes it to form an extra row of somites.

** If there were some way to remove the noggin protein, then they would expect that would cause no somites to form.

The "notch" protein and the "lunatic fringe" protein may also control where and when somites are formed.

Other research papers that report proteins similar to "even skipped"

One theory is based on a combination of a "clock" and a "wavefront"
Some property oscillates up and down, regularly with time. Another property changes steadily.
A combination of these two properties causes cells to form somites, for example, each time the sum of these properties exceeds some threshold, then all the cells that pass the threshold form a somite.

An older theory (by Waddington) was that somite formation is mechanically analogous to the breaking up of a thin stream of water into many individual droplets (as can be seen in a water fountain)

No one asks the question why form somites at all, unless the mechanical process of separation is part of the signal mechanism. If chemical gradients are what is important, why not skip somites?
(i.e. and go directly to dermatomes, myotomes and sclerotomes)

Each somite subdivides into 4 parts:

anterior sclerotome
posterior sclerotome
myotome
dermatome

Myotome cells elongate from front to back. All your voluntary muscles (=skeletal muscles, but not heart muscle or smooth muscles) develop from the myotomes,
including muscles in your arms and legs.

Experiments were done in which myotomes were surgically removed from sections of chick embryos, in which case the legs that developed had no muscles, but only tendons where the muscles should have been.

The following material you will NOT be tested on, but is something to think about.

My own belief (bias/ prediction / hypothesis-I-wish-I-could-test) is that whatever mechanical forces control the geometry of a finished anatomical structure should be more or less the same as the mechanical forces that originally created that structure.

Some specific examples:

For mechanical tension to create muscles: (Muscles exert tension; so tension should create them?)

For mechanical pressure to stimulate formation skeleton:
(Skeleton serves to resist pressure; therefore a good control mechanism to make sure the right amounts of bone gets deposited at each location is for bone cells to detect pressure and deposit more bone at whichever locations "feel" the most pressure)

Cartilage is shaped by a counter-balance between 2 forces:

* Electro-osmotic pressure
* Held back by stretching of collagen fibers.

Notochord depends on a balance between collagen tension and pressure of vacuolated cells;
(Therefore, formation of the notochord should be produced by an imbalance of collagen tension, squeezing cells.)

Much of embryonic development depends on moving cells around. Movement of anything requires some kind of physical forces. Mainline thinking today (as in our textbook) almost totally ignores physical forces. It seems to assume that the noggin protein (just as an example) can cause cells to rearrange into the shape of a somite, and that it's none of our business how the noggin protein causes cells to rearrange or more. In a causal sense, it assumes that mechanical forces are always "downstream" of chemical signals, so much so that the forces are of no interest.