Embryology is a schedule, not a catalogue

The way embryology is first taught makes it look like a catalogue: here are the three germ layers, here is what each one becomes, memorise the table. That framing survives an exam and fails everywhere else. What actually matters is that development is a schedule. Structures appear in a fixed order over a narrow window, each one depending on the one before, and almost every congenital anomaly worth knowing is a fault in the timing rather than a missing item from the list.

Eight weeks. That is the whole embryonic period. By the end of it the body plan is complete and everything afterwards is growth and refinement.

Week one: getting to the wall

Fertilisation happens in the ampulla of the uterine tube. The zygote begins cleaving as it travels, dividing without growing, so the cells get smaller while the whole stays about the same size. By day three or four it is a solid ball — the morula. Then fluid collects inside it and it reorganises into a blastocyst: an outer trophoblast that will become placenta, and an inner cell mass that will become the embryo proper.

That separation, on roughly day five, is the first real decision the conceptus makes. Everything the embryo will be comes from a small eccentric clump of cells; the majority of what has been built so far is destined to be discarded at birth.

Implantation begins around day six, and this is the first point where the schedule can fail visibly. Implantation anywhere other than the uterine body gives an ectopic pregnancy — most often tubal, and dangerous precisely because the tube cannot accommodate what the endometrium can.

Week two: everything comes in twos

Week two is easy to remember because almost everything doubles.

The trophoblast splits into cytotrophoblast and syncytiotrophoblast, the latter being the invasive, multinucleate layer that erodes into maternal vessels and starts producing hCG — which is why a pregnancy test works when it does. The inner cell mass splits into epiblast and hypoblast, forming a bilaminar disc. Two cavities appear: the amniotic cavity above, the primitive yolk sac below.

Nothing recognisable as an embryo exists yet. There is a flat two-layered plate suspended between two bubbles.

Week three: the axis appears

This is the pivotal week. The primitive streak forms at the caudal end of the epiblast, and with it the embryo acquires a midline, a head end and a tail end, a left and a right. Cells of the epiblast migrate through the streak and rearrange themselves into three layers — ectoderm, mesoderm and endoderm. This is gastrulation, and it is the moment the epiblast stops being a sheet and starts being an organism with an architecture.

The notochord extends forward from the streak beneath the ectoderm and signals the overlying cells to thicken into the neural plate. The plate folds, its edges rise and meet, and the neural tube closes — beginning in the middle and zipping towards both ends. The cranial neuropore closes around day 25, the caudal around day 27.

Those two dates carry more clinical weight than almost anything else in the syllabus. Failure of cranial closure gives anencephaly; failure of caudal closure gives spina bifida. Folate supplementation reduces the risk, but only if it is on board before the closure window — which falls before many women know they are pregnant. That is the entire argument for periconceptional supplementation, and it follows directly from the schedule rather than from any fact about folate itself.

The primitive streak normally regresses. When remnants persist, they can give rise to a sacrococcygeal teratoma — a tumour containing tissue from all three germ layers, in the one place where cells capable of making all three lingered too long.

Weeks four to eight: folding and organogenesis

The flat trilaminar disc now folds in two directions at once. Lateral folding wraps the edges ventrally to close the body wall and enclose the gut tube; cephalocaudal folding curls the head and tail towards each other. A flat plate becomes a cylinder with an interior.

Within these five weeks the pharyngeal arches form and contribute to the face, jaw, middle ear and great vessels; the heart tube loops and septates; limb buds appear, elongate and separate into digits; and the gut rotates and returns from its physiological herniation into the umbilical cord.

This is the teratogenic window. Before it, an insult tends to be all-or-nothing — the conceptus either survives intact or is lost. During it, an insult produces structural malformation, and which malformation depends almost entirely on which structure happened to be forming that week. Thalidomide is the standard example, and the reason it produced limb defects so specifically is that the sensitive exposure period coincided with limb bud outgrowth. Give the same drug a fortnight later and the limbs are already made.

After week eight the risk does not disappear, but it changes character: insults in the fetal period impair growth and function, particularly of the brain, rather than rearranging anatomy.

Why the framing matters

Once you hold the calendar rather than the list, a lot of otherwise unrelated facts collapse into one explanation.

A ventricular septal defect and a cleft palate are not conceptually different problems — both are failures of a midline structure to fuse on time. Tracheo-oesophageal fistula is a septation that did not complete. Malrotation is a rotation interrupted partway. Ask when rather than what, and the anomaly tells you roughly which week it happened in.

I am still working out how far this holds. It clearly breaks down for anomalies driven by a single well-characterised gene, where the fault is in the instruction rather than the timing, and I do not yet have a good sense of where the boundary sits. But as a first pass over the embryonic period, treating development as a timetable has been considerably more useful to me than treating it as a table.