The spinosaurus was a dinosaur that lived in the Cretaceous some 170 to 190 million years after dimetrodon, which lived in the Permian long before dinosaurs evolved. So, they are not closely related. But perhaps the spinosaurus independently evolved a hump for shaking its prey to death. It lived in shallow water and, being a theropod, probably lifted it prey out of the water to shake them until their necks snapped. A hump would have been necessary for it to kill prey too large to swallow in one gulp, though with its long snout, it probably also got lots of fish.

Crocodiles also eat prey animals too big for one gulp, but they take them for a death roll to drown them. This works for them because cold-blooded animals can last underwater longer than warm-blooded animals can, especially if the latter panic and start to hyperventilate. That was probably not an option for the spinosaurus because it was a dinosaur killing other dinosaurs.

Most modern lizards eat things smaller than themselves and do not chew - they just swallow. The only large lizard that eats things bigger than itself today is the Komodo dragon, and it kills with venom. So there is no extant carnivore with a hump. But I think the bison is a good proxy for the edaphosaurus and, by extension, we can surmise that the dimetrodon killed by shaking its victims.

I am not a paleontologist and my considerations are more biomechanical. However, I think that in the cases of the herbivorous Edaphosaurus and the carnivorous Dimetrodon, a number of features speak for the display function of their sails. Let me very briefly put down some thoughts on this:

- There are two ways to protect the spines of the dorsal sail from damage: i) high stability, or ii) high bending elasticity.

- In the case of i), the use for exerting strong force is possible, as you discuss for the hump of the American bison and as we proposed for the Spinosaurus swimming underwater.

- At first glance, however, the thin, light spines of Edaphosaurus with cross-bars on the spines speak for ii), i.e., for a light display sail, as also suggested by the comparison with the structure of the crested chameleon's sail.

- I would suggest to consider the cross-bars on the spines as starting points for supporting tendons, which initially gave lateral stability and elasticity to the individual spines and acted similar to the spreaders or crosstrees on the mast of sailing boats. Certainly, they may have played a role in cranio-caudal stabilization of the spines.

- I believe that the cross-bars may have caused problems in the function of possible muscles that should have acted on the side of the spinous processes in the cranio-caudal direction. Note that the carnivorous Dimetrodon had shorter and stronger spines without cross-bars.

- A hump the height of Edaphosaurus' spines would have shifted the animal's center of gravity far upward, potentially destabilizing the animal.

In summary, at least in the case of Edaphosaurus, I am in favor of a display function of the sail.

Jan Gimsa Rostock, 25.02.2023

PS: And, maybe as a funny remark: I do not see anything that speaks against Dimetrodon having poison fangs or a long sticky tongue. :-)

The spinosaurus and the bison seem most similar. I really have no idea what the edaphosaurus' cross bars are for, but that is the principal difference between it and the dimetrodon, which have otherwise very similar anatomy.

For reference, I have attached an image of the modern bison, which is a smaller and gentler version of the long-horned bison that lived during the ice age. With its long curved horns, it clearly fought by twisting from side to side and attempting to hook its attacker.

Images of the extinct animals all seem to be blocked by RG.

On a side note, the triceratops must have formed a phalanx to protect each other's flanks because those long forward-pointing straight horns would have been useless in a melee.

Here is the image of the spinosaurus skeleton, which you can compare to the bison skeleton above.

Incidentally, we don't know that the vertical bones were angled forwards as shown. They could have been angled slightly backwards as the bison's are. The fossilization process leaves the bones in a bit of a jumble.

This is how the spinosauus is typically depicted by artists. Honestly, this image makes no sense. Imagine if paleontologists a million years hence dug up a bison skeleton and drew it with a thin membrane stretched over its vertical bones, perhaps suggesting that it used the wind to literally sail across the plains. Wouldn't that be silly?

There were no grasses in the Paleozoic. The histology of the spines are vastly different from those of extant animals like the bison. The anatomy of the spines and the neural spine proper indicate that axial musculature extended across the veryebral column at the axis which has no elongated spine level of the

Oops, wrote on my cell and things went a bit haywire. You get the point though, i hope. The axis, second vertebra matches the height of the skull and the thickened neural spine portion at the base of each hyperelongated spine. This is where the axial muscles would attach

You're right, Robert Reisz. I was wrong to make this discussion about the Dimetrodon and Edaphosaurus. But I still think that the Spinosaurus, which was a dinosaur and lived 170 to 190 million years after dimetrodon and so was unrelated, had a hump, not a sail. It had a long snout, which was good for catching fish, but it also had the ability to rear up and shake larger prey until their necks broke. Yes?

Incidentally, another animal with thick vertical spines on its back is the wooly rhinoceros. It had a longer nose horn than the modern rhinoceros, and proportionately longer spines on its back. The more an animal depends on twisting side to side to fight (or to plow snow), the longer and thicker the spines on its back. Yes?

Its possible, but to test that hypothesis i would have to examine the specimens and possibly take histological samples of the spines to check such features as the presence and distribution of sharpey’s fibers

Okay. I am a geometer, not a paleontologist, and I don't want to offend you guys by telling you how to do your science. It's just that I had an epiphany: When I saw those silly sail images, I thought, that is an animal who twisted violently from side to side, either to fight (goring attackers like the long-horned bison or the wooly rhinoceros), to kill by breaking necks (like a dog shaking a snake), or to plow snow like the modern bison.

Here is a link to an image of the long-horned bison. You can see that it has spines very similar to the spinosaurus.

https://www.flickr.com/photos/jsjgeology/15444089682/in/photolist-csJwYm-7C5qqn-7C5qqK-pwJZNj-fVs7id-pwLM4k-8TvcvY-fVrZ3T-75NPp7-QZSfNx-RbqWd4-PU5wf7-xh3VvJ

Just to finish the points about Dimetrodon. Rega et al. (2012) performed a nice histological study on pathological Dimetrodon neural spines and found that they were connected by tissue but also subject to large bending stress (lateral deformation) with the "dumbell" cross-section of the spines being important and reestablished quickly after the injury which would not be such an important factor if embedded deep in muscle muscle tissues anchored on them. Additionally the distal tips of the spines show no indication of having been embedded in the sail since they curve irregularly at their distal ends.Article Healed Fractures in the Neural Spines of an Associated Skele...

When you say 'hump', how wide beyond the bounds of the bones are you proposing? 1.2x? 2.0x? 4.0x? Tapering at the top, I presume? At some point, of course, the creature becomes excessively top heavy. On the other hand, if soft tissue remains a traditional sail (e.g. 0.8x to 0.2x neural spine width) then that issue evaporates. Haptodus and Ianthodon are sail/hump-less outgroup taxa for this convergent development. Are transitional taxa informative? I know of a ?juvenile Dimetrodon with relatively much longer legs and smaller sail, Ianthasaurus, Sphenacodon and Echinerpeton.