This appendix is not part of the midrash. It presents observations from comparative anatomy and genomics that create an unexpected parallel to the structure the essay identified in the text. Science does not prove the midrash. The midrash does not depend on science. Both point in the same direction.
Status note: This appendix is an exploratory research notebook. Findings 1–3 are preliminary observations; the SRD5A2 investigation (8 versions) ended in an explicit negative result. The strongest observations are the comparative anatomy and HOXA13 depletion. The remaining findings require repeat masking, control loci, and a phylogenetic model — work not yet performed. This appendix does not carry the central thesis of the essay.
In veterinary anatomy, penile anatomy is commonly described in two broad structural types:
Fibroelastic — bull, ram, boar. The organ is rigid at all times, extruded from an internal sheath like a telescope, and does not change significantly in size. The mechanism is muscular-mechanical. (Hafez & Hafez, Reproduction in Farm Animals, 7th ed.)
Musculocavernous — human, equines (horse, donkey), and also dog — though the dog has an os penis (baculum). Equines share with humans a significant musculocavernous mechanism without a baculum — a combination not found in other domestic animals. The organ is flaccid at rest and engorges via hydraulic blood pressure into spongy chambers. A dramatic change in size and form. (Sisson & Grossman, The Anatomy of the Domestic Animals, 5th ed.; Dyce, Sack & Wensing, Textbook of Veterinary Anatomy)
No internal bone (Baculum) — many mammals (apes, dogs, bears, rodents) carry an internal bone in the organ. Humans and equines — do not. This combination — musculocavernous without a baculum — is not found in bulls, rams, or boars. (Brindle & Opie, Postcopulatory sexual selection, Proc. R. Soc. B, 2016)
Double Invagination — in equines alone, the preputial sheath contains a double fold (preputial ring) that allows dramatic expansion. This structure does not exist in bulls, rams, or boars. (Ashdown & Done, Color Atlas of Veterinary Anatomy: The Horse)
Female Anatomy — in equine females (mare, jenny), the vulvar structure is characterized by prominent vertical labia, a developed clitoris with erectile tissue, and a smooth elastic canal — features not found to the same degree in cows or sows. (Sisson & Grossman; McKinnon et al., Equine Reproduction)
The anatomical bottom line: Among all domestic animals, equines (horse and donkey) share with humans several prominent features in penile anatomy — a hydraulic mechanism, the absence of an internal bone, and a double prepuce — in a combination not found in bulls, rams, or boars. This is an observation of veterinary anatomy, not a claim of the essay.
BovB (Bovine-B) sequences are mobile elements with a well-documented history of horizontal transfer among distantly related animal lineages, including squamates and mammals (Walsh et al., PNAS 2013; Ivancevic et al., Genome Biology 2018).
Important clarification on the numbers: The table below presents counts of all matches (total hits) from RepeatMasker — including fragments, partial matches, and related RTE-family elements. These are not full-length BovB copies. The literature (Ivancevic et al., 2018, Supplementary Tables) reports very few to zero full-length BovB copies in humans and equines. In the cow, by contrast — thousands of full-length copies. The distinction between raw RepeatMasker counts and full-length copy counts is critical, and is presented here for transparency:
| Species | RepeatMasker BovB hits (incl. fragments) | L1 hits | BovB full-length (per Ivancevic) |
|---|---|---|---|
| Human | ~9,120 | ~1,001,412 | ~0 |
| Horse/Donkey | ~11,221 | ~739,445 | <50 |
| Cow | ~568,745 | ~878,559 | thousands of full-length copies |
The similarity in raw fragment counts between human and equine is interesting as a preliminary observation, but does not prove direct horizontal transfer between the two. Phylogenetic analysis of the sequences themselves is required to determine whether these are genuine BovB, related RTE elements, or chance matches. This work is planned.
In the veterinary literature, equines (horse, donkey) are described as having a preputial structure with a double fold (double invagination) — a structure distinct from the prepuce in many other domestic animals, including bulls, rams, and boars, where the structure is single-layered and relatively simple.
In the analysis published in the main book (Tobul, 2026, Chapter 28b), HOXA13 — a developmental regulator involved in genital tubercle patterning and urethral closure — shows extreme depletion of mobile elements:
| Species | Element | HOXA13 enrichment | Genome average |
|---|---|---|---|
| Cow | BovB | ×0.01 | 12.25% |
| Human | L1 | ×0.07 | 17% |
In a genome where L1 comprises 17% of all DNA, the HOXA13 region contains only 7% of average density. Extremely low relative to genome background.
| Layer | Finding |
|---|---|
| Text | "the serpent beguiled me" → donkey → foreskin → circumcision → Zipporah cuts |
| Anatomy | Donkey = musculocavernous + no baculum + double prepuce |
| BovB | Human ~9K fragments ≈ Donkey ~11K (Cow 569K — 50×). Full-length: human ~0, donkey <50, cow: thousands |
| HOXA13 | Developmental regulator: ×0.07 L1 in human, ×0.01 BovB in cow — extreme depletion |
Three observations — comparative anatomy and HOXA13 depletion are the strongest. BovB fragments are intriguing but require validation. Science does not prove that Eve encountered a donkey in the Garden of Eden — that is the essay's imagery, not the language of genomics. But the accumulation is noteworthy: the animal that the Torah pairs with the foreskin, with crouching, and with the sin-offering — is also the animal that shares with humans prominent anatomical features in reproductive regions.
An expanded investigation of SRD5A2 (September 2026) went through 8 analysis versions, all of which failed under adversarial review. There is no donkey-specific finding in coding: donkey = 685/765bp (89.5%), horse = 689/765bp (90.1%) — a difference of 4 bases. All species maintain 88–90%.
The 14bp match at SRD5A2 (a preliminary finding in earlier drafts) — is not evidence of homology (NCBI BLAST FAQ) and has been removed from this appendix.
L1 density differences and unique coverage around foreskin genes — exploratory results requiring repeat masking, control loci, and a phylogenetic model. Moved to a separate research notebook and are not part of this appendix.
The genomic connection requires further work. Future direction: prepuce development gene panel (AR, SHH, FGF10, FGFR2) in regulatory regions; search for structural variant/TE insertion unique to donkey. Not yet performed.