Human embryos develop and lose tails, gill arches, and lanugo

One of the most striking lines of evidence for human evolution comes not from fossils or DNA sequences, but from the development of every human being. During the roughly 38 weeks between conception and birth, a human embryo passes through a series of stages that produce structures inherited from distant evolutionary ancestors—pharyngeal arches homologous to fish gill supports, a tail with true vertebrae, a coat of fine fur, and a yolk sac that contains no yolk.^{1, 2} These structures are not random accidents of development. They reflect deeply conserved genetic programs—particularly the Hox genes—that have been directing animal body plans for more than 500 million years.^{3, 4} The fact that human embryos build, and then dismantle or repurpose, these ancestral structures is among the most compelling demonstrations that humans share common ancestry with other vertebrates.

Pharyngeal arches

During the fourth week of human embryonic development, a series of paired bulges called pharyngeal arches (historically known as branchial or gill arches) appear on either side of the developing pharynx. In fish, these arches develop into the bony or cartilaginous supports for the gills, the organs of aquatic respiration. In human embryos, the same structures form—complete with their own arteries, nerves, and cartilage rods—but are remodeled into entirely different adult structures rather than forming gills.^{1, 5}

The first pharyngeal arch, also called the mandibular arch, divides into maxillary and mandibular processes that give rise to the upper and lower jaws, the palate, and two of the three middle ear bones (the malleus and incus). The second arch, the hyoid arch, produces the stapes bone of the middle ear, the styloid process of the temporal bone, and parts of the hyoid bone in the throat. The third arch contributes to the remainder of the hyoid bone and to the stylopharyngeus muscle. The fourth and sixth arches form the cartilages of the larynx, including the thyroid, cricoid, arytenoid, corniculate, and cuneiform cartilages.^{5, 6}

The pharyngeal pouches—the endodermal linings between the arches—are equally informative. The first pouch becomes the middle ear cavity and Eustachian tube. The second pouch gives rise to the palatine tonsils. The third pouch differentiates into the inferior parathyroid glands and the thymus, a critical organ of the immune system. The fourth pouch forms the superior parathyroid glands and the parafollicular C cells of the thyroid gland.^{6, 7}

The homology between human pharyngeal arches and fish gill arches is not merely a superficial resemblance. The same families of genes—including the Dlx, Hox, and endothelin signaling pathways—pattern these structures in both fish and mammals, and mutations in these genes produce strikingly similar defects across species.^{8, 9} A 2014 review in Wiley Interdisciplinary Reviews: Developmental Biology concluded that the pharyngeal apparatus is "one of the defining features of the chordates" and that its developmental genetic program is "remarkably conserved" across all vertebrate lineages.⁸

Pharyngeal arch derivatives in humans^{5, 6}

The embryonic tail

At approximately four to five weeks of gestation (Carnegie stages 13–14), the human embryo possesses a conspicuous tail extending beyond the developing legs. This is not merely an external protuberance; it contains up to 13 to 16 developing caudal somites, a continuation of the spinal cord (the caudal neural tube), blood vessels, and mesenchymal tissue.^{10, 11} The tail reaches its maximum length relative to the embryo at Carnegie stage 16, when the number of caudal somites peaks. By Carnegie stage 23 (approximately eight weeks), the tail has been almost entirely eliminated, leaving behind only the three to five fused vertebrae of the adult coccyx.¹⁰

The mechanism of tail removal is programmed cell death, or apoptosis. A 2018 study by Tojima and colleagues in the Journal of Anatomy used serial histological sections of human embryos from the Kyoto Collection to map the process in detail. They found that the number of caudal somites decreased dramatically after Carnegie stage 16, with approximately five somites disappearing through apoptosis. The tail's regression followed a proximal-to-distal pattern, with the most distal vertebrae vanishing first.¹⁰ Earlier work by Fallon and Simandl in 1978 had established that cell death is the primary mechanism responsible for the disappearance of the human embryonic tail, demonstrating extensive apoptosis in the ventral mesoderm and regressing tailgut.¹²

A 2024 study published in eLife by Chedotal and colleagues provided the most detailed three-dimensional reconstruction of tail development to date, confirming that the human embryonic tail contains genuine vertebral elements and a fully formed caudal neural tube that are progressively dismantled by apoptosis.¹³ The genetic pathways controlling tail regression appear to be conserved across mammals: downregulation of the Wnt-3a signaling pathway triggers apoptosis in mouse tail cells, and disruption of this pathway in mice results in tail retention.¹⁴

Babies born with tails

In rare cases, the apoptotic program that normally eliminates the embryonic tail fails to complete, and a baby is born with a caudal appendage. A landmark 1982 case report in the New England Journal of Medicine by Fred Ledley documented a newborn with a well-formed tail-like structure and explicitly framed the finding as evidence for human evolution.¹⁵ A 2016 review by Tubbs and colleagues in Clinical Anatomy cataloged the known cases and classified human tails into "true tails" and "pseudotails." True tails contain adipose tissue, connective tissue, striated muscle bundles, blood vessels, nerve fibers, and Vater–Pacini sensory corpuscles, all covered by normal skin with hair follicles and sweat glands.¹⁶

Fewer than 60 well-documented cases of human caudal appendages have been reported in the medical literature, with approximately half associated with underlying spinal anomalies such as spina bifida occulta.^{16, 17} True tails, which lack bone and cartilage but contain voluntary muscle and can sometimes move, are interpreted as atavisms—the reappearance of an ancestral trait that is normally suppressed during development. Their existence demonstrates that the genetic instructions for building a tail remain present in the human genome, silenced but not lost.^{15, 16}

Lanugo

By approximately the sixteenth week of gestation, the human fetus begins to grow a coat of fine, soft, usually unpigmented hair called lanugo. By week 20, this hair covers nearly the entire body, giving the fetus a fur-like coating. Lanugo is the first hair produced by fetal hair follicles and is distinct from the vellus hair (fine body hair) and terminal hair (scalp, eyebrow) that replace it later.^{2, 18}

The fetus sheds most of its lanugo between weeks 33 and 36 of gestation. The shed hair enters the amniotic fluid and is eventually swallowed by the fetus, becoming a component of meconium—the dark, tar-like substance that constitutes the newborn's first stool. Premature infants are often born with visible lanugo still present on their skin, particularly on the shoulders, back, and forehead.¹⁸

Lanugo serves a functional role during fetal development: it helps bind the vernix caseosa, a waxy coating that protects fetal skin from the constant immersion in amniotic fluid. The hair also stimulates sensory receptors in the skin, and this tactile stimulation appears to promote fetal growth during mid-gestation.¹⁸ Nevertheless, the very existence of a full-body fur coat in a species that is functionally hairless at birth is most parsimoniously explained as an evolutionary inheritance. Other primates, including monkeys and apes, develop a similar fetal hair coat at a comparable stage of development, and in furred species this coat simply continues growing rather than being shed. The human pattern—grow a coat of fur, then discard it before birth—reflects the ancestral mammalian developmental program, modified by the evolutionary reduction of body hair in the human lineage.^{2, 19}

The yolk sac

Among the earliest structures to form in the human embryo is the yolk sac, a membranous pouch connected to the developing gut. In birds and reptiles, the yolk sac encloses a large reservoir of yolk that provides nutrition throughout embryonic development. In placental mammals like humans, the yolk sac contains no yolk whatsoever—the embryo receives its nutrition from the mother via the placenta.^{20, 21}

The human yolk sac is not, however, a functionless vestige. It has been repurposed to serve several critical roles during the earliest weeks of development, before the placenta is fully established. The yolk sac is the site of the embryo's first blood cell production (primitive hematopoiesis), generating the initial population of red blood cells and macrophages as early as three to four weeks after conception. It also produces primordial germ cells, which later migrate to the developing gonads to become eggs or sperm. Additionally, the yolk sac synthesizes proteins involved in nutrient transport, lipid metabolism, and blood clotting.^{20, 21}

A 2020 study in Nature Communications by Zhu and colleagues used single-cell transcriptomics to characterize the functions of the primate yolk sac in unprecedented detail, confirming its roles in hematopoiesis, germ cell specification, and metabolic support. The authors noted that while the yolk sac has been "functionally repurposed" in mammals, its basic structure and developmental origin are clearly homologous to the yolk-containing sac of reptiles and birds—a structure that makes functional sense only when viewed through the lens of evolutionary descent from egg-laying ancestors.²¹

Von Baer and the hourglass model

The observation that embryos of different vertebrate species resemble one another more closely than the adults do has a long history in biology. In 1828, the Estonian embryologist Karl Ernst von Baer formulated four laws of embryology based on his comparative observations of vertebrate development. Von Baer noted that the most general characters of a large group of animals appear earlier in the embryo than the more specialized characters; that development proceeds from the general to the specific; and that embryos of different species within a group resemble one another more than the adults do.^{22, 23}

Von Baer's observations were not framed in evolutionary terms—he was skeptical of transmutation theories—but Charles Darwin recognized their profound implications. In On the Origin of Species (1859), Darwin argued that embryological resemblances among vertebrates are best explained by common descent: if fish, reptiles, birds, and mammals all share a distant ancestor, then their embryos should retain the developmental programs inherited from that ancestor, with species-specific modifications appearing later in development.²³

Modern developmental biology has largely vindicated a refined version of von Baer's observations, known as the "developmental hourglass" model. This model proposes that vertebrate embryos are most similar to one another at a mid-developmental stage called the phylotypic period—the stage at which the basic body plan (the "zootype") is established—while diverging more in both earlier and later development. A 2010 study by Kalinka and colleagues in Nature provided quantitative support for the hourglass model by comparing gene expression across six Drosophila species, and a parallel study by Domazet-Lošo and Tautz showed a similar pattern in zebrafish, with the phylotypic stage showing the oldest average gene age.^{24, 25}

Developmental hourglass: gene expression divergence across embryonic stages^{24, 25}

Hox genes and conserved pathways

The molecular explanation for why human embryos pass through these ancestral stages lies in the extraordinary conservation of developmental genes across the animal kingdom. The most famous of these are the Hox genes—a family of transcription factors that specify the identity of body segments along the head-to-tail axis. Hox genes were first discovered in the fruit fly Drosophila melanogaster in the 1980s, when Edward Lewis, Christiane Nüsslein-Volhard, and Eric Wieschaus demonstrated that mutations in these genes cause homeotic transformations, in which one body segment develops the identity of another.³

The subsequent discovery that essentially the same Hox genes exist in vertebrates—and are arranged in the same order along the chromosome, activated in the same sequence during development, and expressed in the same relative positions along the body axis—was one of the most important findings in twentieth-century biology. This property, known as collinearity, was characterized by Denis Duboule and colleagues, who showed that the physical order of Hox genes on the chromosome corresponds to both their spatial expression along the anterior-posterior axis and their temporal order of activation during embryonic development.^{4, 26}

Humans possess 39 Hox genes arranged in four clusters (HOXA, HOXB, HOXC, and HOXD) on four different chromosomes. These clusters are directly homologous to the single Hox cluster found in insects and the duplicated clusters found in other vertebrates. The conservation is remarkable: human and fly Hox genes can sometimes substitute for each other across more than 500 million years of evolutionary divergence. When a mouse Hox gene is experimentally introduced into a Drosophila embryo, it can partially rescue the function of the corresponding fly gene.^{3, 27}

This deep conservation explains why human embryos develop pharyngeal arches, tails, and other ancestral structures. The genetic toolkit that builds a vertebrate body plan has been inherited, with modifications, from a common ancestor. The pharyngeal arches form because the same signaling pathways (including Hox, Dlx, and endothelin) that build gill supports in fish are activated in mammalian embryos; they simply receive different downstream instructions that redirect development toward jaws, ear bones, and laryngeal cartilages rather than gills.^{8, 9} The embryonic tail forms because the same Wnt and Hox signaling cascades that produce tails in tailed mammals are activated in human embryos; the tail is then removed by an additional program of apoptosis that is absent or reduced in species that retain tails.^{10, 14}

Evolutionary developmental biology

The integration of evolutionary biology with developmental biology—a synthesis known as "evo-devo"—has transformed our understanding of how body plans evolve. The central insight of evo-devo is that evolution works largely by modifying existing developmental programs rather than inventing new ones from scratch. Changes in when, where, and how much a gene is expressed (rather than changes in the gene's protein-coding sequence) are often responsible for major differences in body form between species.^{27, 28}

Sean B. Carroll, one of the founders of the evo-devo field, has argued that the conservation of developmental toolkit genes across the animal kingdom constitutes one of the strongest lines of evidence for common descent. In his influential 2005 book Endless Forms Most Beautiful and the accompanying scientific literature, Carroll documented how the same signaling pathways—including Hox, Hedgehog, Wnt, BMP, and Notch—are used repeatedly to build different structures in different organisms, from insect wings to vertebrate limbs to the eyespots on butterfly wings.²⁸

The human embryological features discussed in this article are textbook examples of this principle. The pharyngeal arches, the embryonic tail, the lanugo coat, and the yolk sac are not accidental or meaningless developmental detours. They are the visible consequences of a vertebrate body-building program that is hundreds of millions of years old, one that human embryos execute faithfully before overlaying the modifications that produce a distinctly human body.^{1, 2, 27}

Significance for common descent

The developmental evidence presented here converges with evidence from comparative anatomy, the fossil record, and molecular genetics to support the conclusion that humans share common ancestry with other vertebrates. Each line of embryological evidence is individually suggestive; together they are compelling. The pharyngeal arches demonstrate that humans develop gill-arch homologs controlled by the same genes that build gills in fish.^{5, 8} The embryonic tail demonstrates that humans retain the genetic capacity to build a post-anal tail with true vertebrae, a defining feature of chordates that is actively dismantled by apoptosis.^{10, 13} Lanugo demonstrates that human fetuses execute the ancestral mammalian program of growing a fur coat, then discard it in a species-specific modification.^{2, 18} The yolk sac demonstrates that human embryos build a structure whose original function—enclosing yolk—was lost when the mammalian lineage evolved placental nutrition, but which has been retained because it acquired new functions.^{20, 21}

None of these observations makes sense if each species were independently designed. A designer starting from scratch would have no reason to equip a human embryo with gill arches, a tail, a fur coat, and a yolkless yolk sac only to remove or repurpose them. These features make sense only as the inherited developmental legacy of ancestors for whom each structure served its original purpose—ancestors who breathed through gills, balanced with tails, insulated with fur, and nourished their young with yolk.^{1, 2, 27}