The humerus (or the anterior propodium) is one of the most complex bones of the limb, next to the femur (or posterior propodium), and various carpal/tarsal bones as they are described in mammals (although that’s largely practical, rather than actual). I don’t come to this conclusion lightly, but based on the weight of the number of different features associated with it:
First, the humerus contacts at least four other bones (coracoid, scapula, radius, and ulna) in virtually all tetrapods.
Second, the humerus is encased, liberally, in tendons, ligaments and muscles. Each of these leaves a mark on the humerus as it has its origin on it, or inserts onto it.
Thirdly, depending on the anatomy, the humerus is either a weight-bearing bone in quadrupeds and is therefore the primary bone which produces propulsive thrust during the walk/run cycle, or it is a food-acquisitive element and must have the leverage to act against a variety of contradictory forces or the agility to move the limb in often quite complex ways, or the limb is a thrust producing bone in another way, providing strength to the wing-flap in bats, pterosaurs, and birds.
This is not to say there are no other complex bones in the anterior limb, but for the most part it is the humerus that draws the most attention because it, unlike the other bones of the anterior limb, is also the largest and most prominent. It is not always the longest, but it is almost certainly always the largest, in area, mass, and prominence.
Here is a diagram I produced (slightly outdated as it lacks a “caudal” view), with the abbreviations below, of a standard therizinosauroid humerus, based not lightly on that of Erlikosaurus andrewsi, GIN 100/111 .
add. – adductor crest
cap.h. – humeral crest
cb.fos. – fossa for the m. coracobrachialis
dist.fos. – distal fossa
dist.lam. – distal lamina of the deltopectoral crest
dpc. – deltopectoral crest
ect. – ectepicondylar crest
ent. – entepicondylar crest
ext. – extensor crest
intcond. – intercondylar groove
prox.fos. – proximal fossa (for the m. pectoralis)
prox.med. – proximomedial crest
rc. – radial condyle
sup.cond. – supracondylar ridge
sup.ect. – supraectepicondylar crest
sup.ent. – supraentepicondylar crest
tric.c. – crest and fossa for the m. triceps
uc. – ulnar condyle
There are many flaws: The labeling is messier than it should be, I am not using the terminology of Baumel and Witmer , and I am not even fully describing the features shown. Most of all, I am not even showing a caudal view, which has further to say on the anatomy of a humerus. But this diagram gives you an idea of where things are in general on the humerus. A major note, though, the “cb.fos.” would almost certainly be better called the “pf” for “pectoralis fossa,” but it may be differentiated from the fossa on the medial surface of the deltopectoral crest, even if they generally support the same muscle. I was young, naïve, and stupid when I wrote that caption (well over ten years ago!); and I’ve chosen to leave it this way for the moment, until I fix up a new version.
Variation in the humerus is more complex, perhaps, due to the various features and functions that it can be exapted for, than other limb bones with the exception of portions of the manus. The deltopectoral crest’s shape is relates to at least three muscles (the mm. coracobrachialis, pectoralis, and deltoideus, the latter two for which it is named for), and it will often change orientation and position relating to these muscles, their size, and the humerus’ own size. Based on the relationship and potential mobility of the forearm, the distal condyles of the humerus also vary widely, as does the shape of the humeral shaft in various perspectives. The humeral shaft’s cross-sectional shape varies, not just when the orientation of a crest changes or a ridge extends into that particular region being examined, but through the simple perspective that some humeri are “flatter” than others, or have a slight lenticularity (where the anterior fossa of the humerus extends completely down the shaft, as it does in several therizinosauroids). Moreover, even without changing the shape of the distal condyles, they may alter their position on the humerus, being either distal, proximal, or extending onto the caudal surface of the bone; they may be very close together, conjoined, or widely separated; they may produce extra ridges, or extra subcondyles; and of course, they may be spherical, oblate, cylindrical, or even flat or pyramidal.
The humerus is a fascinating bone in its complexity, both due to size, relationship to bones and muscles, and its roles and function in locomotion and food-gathering. The variation in the humerus is under-appreciated by many, and largely touched on when some works explicitly analyze it for some taxa, but otherwise leave this variation out of the broad picture.
As an aside, I’ve chosen the therizinosauroid humerus because it illustrates two things: 1, therizinosauroid humeri exhibit remarkable variation for seeming to reflect a small, canalized morphotype in many other features of the skeleton, and 2, the therizinosauroid humerus exhibits a remarkable plethora of additional features found all together in a single element that are normally expressed individually in other taxa. It makes it an idea morphology to describe all of the enormous variety present in not just theropod dinosaurs, but tetrapods in general.
 Barsbold R. & Perle A. 1980. Segnosauria, a new infraorder of carnivorous dinosaurs. Acta Palaeontologica Polonica 25(2):187-195.
 Baumel, J. J. & Witmer, L. M. 1993. Osteologia. pp. 45–132 in Baumel (ed.) Handbook of Avian Anatomy: Nomina Anatomica Avium, Second Edition. Publications of the Nuttall Ornithological Club 23. (Cambridge.)