Cut an onion and your eyes stream within seconds. Sit with bad news and your eyes also stream, but the experience is nothing like the onion. One leaves you clear-eyed in a minute, the other leaves you with a headache, a blocked nose and eyelids that are visibly swollen the next morning. I used to assume this was purely about intensity — more tears, more swelling. It is not. The two fluids are measurably different, and the difference tells you what each one is for.

Three tears, one plumbing system
The eye runs a continuous irrigation system. Above each eye, tucked into the orbit, sits the lacrimal gland, which produces the aqueous layer of the tear film. Under the eyelid, small accessory glands add mucus; along the lid margins, the meibomian glands add an oil layer that slows evaporation. Every blink spreads the film across the cornea, and the excess drains through a small opening at the inner corner into the nose — which is why crying hard makes your nose run.
Physiologists usually divide tears into three kinds. Basal tears keep the cornea wet and are always present, whether you notice them or not. Reflex tears respond to an irritant — onion vapour, wind, a bit of grit — and exist to flush it out. Emotional tears appear with grief, relief, rage or a wedding speech, and they are the odd one out. Baboons and macaques have lacrimal glands and produce basal and reflex tears, but as far as anyone has documented, they do not weep from feeling. Something about the human arrangement is different, and the difference is not in the plumbing.
The chemistry that separates them
The cleanest evidence comes from work done in the 1980s by the biochemist William H. Frey II at the University of Minnesota. Frey collected reflex tears from people slicing onions and emotional tears from people watching something sad, then compared the two. The emotional tears carried 20 to 25 percent more protein by type than the reflex ones. They were also far richer in potassium — roughly four times the concentration — and in manganese, at around thirty times the level found in blood serum.
More intriguing was the hormonal content. Emotional tears contain adrenocorticotropic hormone, ACTH, which the body releases under stress, and prolactin, the hormone that also controls the neurotransmitter receptors involved in tear release. Frey also reported leucine-enkephalin, an endogenous opioid fragment associated with pain relief. None of these appear in anything like the same concentration in onion tears.
The composition tells you that emotional tears are not simply more of the same fluid. They are a different secretion, produced under different control, carrying molecules that the body uses to signal stress. Frey's interpretation was that weeping is a way of shedding the chemical products of intense emotion — an idea that became the popular "cry it out" advice.
What the evidence does not support
Here the story needs a careful brake. The detoxification claim is intuitive and widely repeated, and it is not established. As Scientific American's account of Frey's work notes, no study has shown that enough hormone leaves the body in tears to change how you feel, and the relief people report after crying has other plausible explanations: the social signal, the physical exhaustion, the shift in attention. The 20 to 25 percent protein figure is solid. The conclusion that crying removes stress hormones in a meaningful quantity is a hypothesis that has not held up.
What does hold up is the osmotic explanation for the swelling. Emotional tears carry more protein and more potassium than reflex tears, which raises the solute concentration of the fluid sitting on the eyelid and in the surrounding tissue. Water follows solute. The loose tissue around the eye takes on fluid, and by morning the lids look heavy. The puffiness is a small piece of evidence that the fluid really is different, visible in the mirror rather than in a lab.
A cornea without blood vessels
There is a structural reason the eye needs such an elaborate tear system, and it is the part I find most elegant. The cornea has to stay transparent, and blood vessels would scatter light. So the cornea is avascular: no capillaries deliver oxygen, nutrients or immune cells, and no vessels carry away waste. It is fed instead by the tear film on the outside and the aqueous humour on the inside, and everything else arrives by diffusion. Oxygen has to come from the air through the film, which is why a contact lens that fits badly can starve the surface it sits on.
That design puts enormous weight on the tear film. The lipid layer from the meibomian glands has to be intact, or the aqueous layer evaporates and the cornea dries. The mucin layer has to be present, or the water does not spread evenly over a surface that repels it. When any layer fails, the result is dry eye: a surface that cannot maintain its film, exposed to air, with no blood supply to fall back on. It is a reminder that the eye's transparency is bought with a permanent maintenance job.
What the microscope shows
The lacrimal gland looks like a compressed bunch of grapes in section: clusters of secretory acini, each a ring of cells around a small lumen, draining into ducts. The cells are packed with secretory vesicles, and the whole organ is embedded in the connective tissue of the orbit. Students who have looked at salivary gland or pancreas sections recognise the architecture immediately, because it is the same problem — make a fluid, deliver it through a duct.
In the conjunctiva, the thin membrane lining the inside of the lid, the cells you want to find are goblet cells: pale, flask-shaped cells wedged among the epithelial cells, full of mucin. In a stained section they stand out as clear vacuoles against the darker epithelium. Their number and distribution matter clinically, because they are the source of the layer that keeps tears from beading off the cornea.
And the cornea itself is one of the best histology specimens in any syllabus: five layers, a thick regular stroma of collagen lamellae, and no vessels at all. Once you have seen that, the tear film stops being a detail. It is the only blood supply, in effect, that the front of the eye has.
Why this belongs in a biology lesson
Tears are a good case for teaching that a fluid can carry information about its function. The same gland produces three secretions on three different triggers, and the composition changes with the trigger.
It is also a good case for teaching caution. Frey's protein data are replicated and cited everywhere; his detox hypothesis is not, and the popular version of his work has drifted well past what the studies show. Being able to separate the two — a measured difference versus an appealing interpretation — is the skill that most of the biology syllabus is really trying to build.
If you would rather examine the tissue than take my word for the section, Wwai's scanned specimens and AI lab partner let you work through lacrimal gland, conjunctiva and cornea at your own pace.