# Blood Vessels, Blood and Lymph

> Fluid flows down a pressure difference, at a rate set by the resistance in the way; across a capillary wall, the difference that counts is blood's push outward minus the plasma proteins' pull inward. This lesson uses that rule to show why arterioles, not the narrower capillaries, hold most of the resistance and why blood is slowest in the capillaries; works the Starling forces through with numbers, from Landis's frog capillaries to the revised view in which lymph, rather than steady reabsorption at the venous end, returns filtered fluid in most tissues; and lays out what blood contains and how it stops a leak. Afterwards you can calculate a resistance, predict the effect of narrowing an arteriole, and judge what the evidence on capillary exchange does and does not show.

Course: Organismal Biology (Supply and exchange). Level: first- and second-year university. About 19 minutes.
Lesson page (link this one when citing): https://stemhop.com/lessons/blood-and-vessels/

Builds on: Circulation and the Heart · Size, Scale and Surface · Tissues, Organs and Homeostasis

## What you'll learn

- Relate the wall of each vessel class (artery, arteriole, capillary, vein) to its job, and define compliance.
- Use flow = pressure difference / resistance and Poiseuille's 1/r^4 idealization to calculate resistances and flow changes, and explain why blood is slowest in capillaries.
- Calculate the net Starling pressure across a capillary, interpret Landis's frog mesentery experiment, and explain why lymph is needed to return filtered fluid and protein.
- Name the components of blood and the steps of hemostasis, and compare them in mammals, other vertebrates and the horseshoe crab.

**The first film showed why blood is slowest in the narrowest vessels: the same flow spreads over the largest total area.** It left a second puzzle: capillary walls leak water, so why doesn’t the plasma drain into the tissues? One rule answers both.

> **The one idea.** Fluid flows down a pressure difference, and how much flows is that difference divided by the resistance in the way: Q = (Pin − Pout) / R. Across a capillary wall the driving pressure is the blood’s push outward minus the plasma proteins’ pull inward, each a difference between inside and outside the wall.
>
> *Scope:* steady flow through a simplified vessel tree at one height, with no pump inside and no blood bypassing the capillaries (gravity, pumps and shunts add terms of their own), and fluid crossing capillary walls; not the diffusion of gases and solutes, the pulse, or what blood is made of. Later mentions of the rule assume these conditions rather than repeat them.

## Each wall fits its job

In a closed circulation (see [circulation and the heart](https://stemhop.com/lessons/circulation-and-heart/)) blood leaves the heart in **arteries** and returns in **veins**. The names describe direction, not oxygen: in mammals the pulmonary artery carries oxygen-poor blood away from the heart, and the pulmonary veins carry oxygen-rich blood toward it. Between them lie the **arterioles**, the smallest arteries, and the **capillaries**, each with a wall fit for its work ([tissues and homeostasis](https://stemhop.com/lessons/tissues-and-homeostasis/)).

- **Arteries** have thick walls with smooth muscle and elastic fibres. Their elastic recoil between beats keeps pushing blood forward, so arterial pressure never falls to zero.
- **Arterioles** have walls dominated by smooth muscle. When it contracts (**vasoconstriction**) the arteriole narrows; when it relaxes (**vasodilation**) it widens. These are the film’s taps, holding most of the circulation’s resistance. Besides nerves ([the hypothalamus and autonomic system](https://stemhop.com/lessons/hypothalamus-autonomic/)) and hormones, local signals set their tone: stretch of the wall, chemicals from active tissue, and nitric oxide from the endothelium.
- **Capillaries** have a wall of a single layer of **endothelial cells**, the flat cells that line every vessel, on a thin basement membrane. Little else separates blood from tissue fluid, so exchange happens here (how the cells join: [epithelia](https://stemhop.com/lessons/epithelia/), [adhesion and the extracellular matrix](https://stemhop.com/lessons/adhesion-and-ecm/)).
- **Veins** have thin walls and wide channels. They are more **compliant** than arteries: **compliance** is the change in volume per change in pressure, so they hold much more blood for a small rise in pressure. In adult humans at rest, most of the blood is in the veins. Many limb veins have one-way valves; contracting muscles squeeze them and push blood toward the heart: the **muscle pump**.

*Figure (AI-generated illustration).* Each wall fits its job: arteries recoil to keep pressure up between beats, arterioles set resistance with smooth muscle, capillaries exchange across one layer of endothelial cells, and compliant veins hold most of the blood, with valves and muscles pushing it toward the heart. Image: four flat schematic vessel segments in a row. A thick-walled artery with muscle and elastic fibres, a narrow muscular arteriole, a capillary with a one-cell wall, and a wide thin-walled vein with valves. An inset shows a vein with valves being squeezed by muscle, pushing blood toward the heart.

## Numbers on the 1/r4 lever

For steady flow of a simple fluid through a rigid tube, **Poiseuille’s law** gives resistance from the fluid’s **viscosity** (its thickness: honey resists more than water), the tube’s length and its radius:

> **R = 8 × viscosity × length / (π × radius^4)**

Blood is not a simple fluid and vessels are not rigid, so this is an idealization.

**Worked problem.** Illustrative numbers for an adult human at rest: mean arterial pressure (the average pressure in the aorta) 93 mmHg, right atrial pressure (where blood re-enters the heart) 3 mmHg, cardiac output (the heart’s flow) 5.0 L per min.

1. *The whole body as one resistance.* Total peripheral resistance = (93 − 3) / 5.0 = 90 / 5.0 = 18 mmHg min per L. That lumps every vessel together.
2. *One arteriole narrows.* Its radius halves, all else fixed. Its resistance rises 24 = 16 times, so its flow falls to 1/16, or 6.25 percent. If you got one half, you treated resistance as proportional to 1/r.

**Predict.** Now every arteriole in one organ halves its radius. Resistances in a chain (**in series**) add. If the arterioles held half the organ’s resistance and the pressure across the organ is fixed, does the organ’s flow also fall to 1/16?

> **Check your prediction.** No. The arterioles’ half becomes 0.5 × 16 = 8 and the rest stays 0.5, so the organ’s resistance becomes 8.5 times the old and its flow falls to 1/8.5, about 12 percent. One sixteenth describes the narrowed vessels, not the organ. Narrowing a single arteriole among many side by side would change the organ’s flow very little.

### Narrowest, but not the main resistance

Paths side by side (**in parallel**) work the other way: the flow divides, so n equal paths resist 1/n as much as one. With a made-up figure, if one capillary resisted 50 times as much as one arteriole, 100 side by side would resist half as much. Real beds also branch and rejoin, so this sketches the logic only.

Behind the film’s speeds, in made-up round numbers: 5 L per min of **volume flow** (how much passes a point) is about 83 cm3 per s; 83 / 4 cm2 of aorta is about 21 cm per s, and 83 / 3,000 cm2 of capillaries is about 0.03 cm per s.

Slow blood in a leaky tube brings back the second puzzle, and the second film follows the first measurement of fluid crossing one capillary’s wall.

**Film transcript: Landis's Blocked Capillary**

### Film transcript: Landis's Blocked Capillary

Capillary walls let water and small solutes through. So why doesn't all your plasma end up in your tissues? Something must hold most of it in. The first film showed that blood flows down a pressure difference, divided by a resistance. Across the capillary wall, the same rule decides which way fluid moves. Two pressures act across the wall. The first is blood pressure inside the capillary, minus the tissue fluid's pressure outside. It pushes fluid outward. Physiologists call it hydrostatic pressure. The second comes from plasma proteins, mostly albumin. Few cross the wall, so they pull water back in. Protein outside pulls the other way, so the difference counts.

That pull is the oncotic pressure. In 1896, Ernest Starling, working with dogs, argued that these two pressures set the direction of flow. But neither had been measured inside a single capillary. In 1927, Eugene Landis took up the test in frogs. He spread out the mesentery, the thin sheet that holds the gut, and watched single capillaries under a microscope. A fine glass pipette slid into one capillary to record its pressure. Then he pressed the capillary shut with a fine glass rod. With the flow stopped, the trapped red cells drifted slowly. They moved toward the block when fluid was leaving through the wall. They moved away when fluid entered.

From the cells' speed and the vessel's size, he worked out the fluid crossing each patch of wall. Then he plotted that rate against the pressure he had measured. If Starling was right, what should happen when capillary pressure falls below the pull of the plasma proteins? Which way should the fluid go?

Fluid should come in, and it did. Filtration rose roughly in proportion to pressure. It fell to zero near the oncotic pressure of frog plasma, then reversed. So across the wall, the driving pressure is the push minus the pull. The more the push wins, the faster fluid leaves. When the pull wins, fluid comes in. But Landis measured single capillaries in one frog tissue. His result ruled in the balance there. It did not show what every capillary does over hours. Later reviews concluded that a lasting return at the venous end lacks reproducible support. In most tissues, capillaries filter slowly, and lymph vessels carry the fluid back.

From the film

- Across a capillary wall, the driving pressure is the blood's push outward minus the plasma proteins' pull inward, each a difference between inside and outside the wall.
- Landis blocked single frog capillaries and timed the drift of trapped red cells to measure fluid crossing the wall.
- Filtration rose roughly in proportion to capillary pressure and reversed near the oncotic pressure of frog plasma.
- His experiment did not show that fluid keeps returning at the venous end; in most tissues, lymph returns it.

## Starling’s balance in numbers

The film weighed push against pull without sizes. In symbols, with P for **hydrostatic pressure** (the pressure of the fluid itself), π for **oncotic pressure** (the osmotic pull of proteins that cross the wall poorly, in mammals mostly albumin), c for capillary, i for **interstitial** (the fluid between cells) and outward positive:

> **net pressure = (Pc − Pi) − (πc − πi)**

This form assumes protein cannot cross the wall at all (Going deeper gives the full relation). Flow across the wall is this net pressure divided by the wall’s resistance to water: the one idea again.

*Figure (AI-generated illustration).* Trapped red cells work as a flow meter: their drift gives the rate of filtration at the pressure the pipette records. Image: a thin transparent sheet of frog mesentery beside a loop of gut, with one capillary pierced from above by a fine glass pipette and pinched shut at one end by a glass rod. Two insets show red cells in the blocked segment drifting toward the block when fluid leaves the wall and away from it when fluid enters.

### The classic model, worked

Illustrative adult human numbers: capillary pressure 35 mmHg at the arterial end, falling to 15 mmHg at the venous end as pressure drops along the capillary’s own resistance; interstitial pressure 0; plasma oncotic pressure 25 mmHg; interstitial oncotic pressure 3 mmHg. The oncotic difference is 25 − 3 = 22 mmHg. Arterial end: 35 − 22 = +13 mmHg, outward. Venous end: 15 − 22 = −7 mmHg, inward.

*Interactive on the lesson page:* Starling forces along a capillary

### What survives, and what does not

The film’s verdict: the outward push survives, but a steady venous-end return does not hold as a general account for most tissues. The reason is the oncotic term. The difference that matters lies across the **glycocalyx**, a sugar-rich coat on the inner surface of the endothelium that holds back protein: between plasma and the fluid just beneath the coat (πsub, which replaces πi in the equation). That fluid has less protein than average tissue fluid, so tissue protein does not cancel the pull. In perfused rat mesenteric microvessels (fed with a fluid of set make-up), the effective oncotic opposition stayed near 70 percent of the plasma oncotic pressure inside the vessel, even with equal albumin in plasma and tissue (Adamson and colleagues, 2004). Applied, for illustration only, to 25 mmHg: 17.5 mmHg.

**Predict.** With that 17.5 mmHg pull and interstitial pressure 0, which way does fluid move at a capillary pressure of 35 mmHg, and at 15 mmHg?

> **Check your prediction.** At 35 mmHg, net pressure is 35 − 17.5 = +17.5 mmHg, outward. At 15 mmHg it is 15 − 17.5 = −2.5 mmHg, inward, but only for a while. Water drawn in leaves the protein beneath the coat more concentrated, the pull shrinks, and the absorption fades.

Some beds do absorb steadily: those where an epithelium, a sheet of lining cells, keeps supplying fluid for the capillaries to take up.

The giraffe stretches these forces: its feet are far below its heart, and its arterial pressure is much higher than ours. Hargens and colleagues (1987) reported three defences against fluid build-up in its legs: a tight skin layer that, with pressure changes during exercise, helps move fluid upward against gravity; narrowing of the small vessels that feed the capillaries; and capillary walls that let little plasma protein through.

## Lymph: the return route

Filtered fluid enters blind-ended **lymphatic capillaries** and is then called **lymph**. In mouse trachea their cells are joined by short “buttons” with loose, overlapping flaps between them, thought (from structure alone) to let fluid in but not out; larger collecting vessels have continuous “zipper” junctions and valves.

**Predict.** Frogs and other amphibians have **lymph hearts**, small pulsating sacs that pump lymph into veins. Mammals have none. What moves theirs?

> **Check your prediction.** The walls of mammalian collecting lymphatics contract, body movements squeeze them, and their valves keep lymph moving one way, much as the muscle pump moves venous blood.

Lymph passes through nodes, where white cells sample it (see [innate immunity](https://stemhop.com/lessons/innate-immunity/)), and in humans it re-enters the blood where the jugular and subclavian veins join in the neck. Leaked protein comes back mainly this way. In the small intestine, lymphatics called lacteals carry absorbed fat (see [digestive systems](https://stemhop.com/lessons/digestive-systems/)).

## Blood and stopping a leak, compared across animals

A cut opens a new path out of the vessel; the blood carries what closes it.

Blood is **plasma**, the liquid, carrying cells. Human plasma is more than 90 percent water, with proteins including albumin, globulins and fibrinogen. **Red cells** (erythrocytes) carry hemoglobin, the oxygen-binding protein (see [oxygen transport](https://stemhop.com/lessons/oxygen-transport/)). **White cells** (leukocytes) defend the body. **Platelets** are cell fragments shed by large cells called megakaryocytes. The fraction of blood volume made up of red cells is the **hematocrit** (Going deeper gives the human range).

Mammalian red cells lose their nucleus as they mature; those of birds, reptiles, amphibians and fish keep it, with rare exceptions such as a few salamanders. Platelets are a mammal feature: other vertebrates have nucleated **thrombocytes** that do a similar job.

*Figure (AI-generated illustration).* Mammals have red cells without a nucleus and platelets; almost all other vertebrates keep the red-cell nucleus and use nucleated thrombocytes instead of platelets. Image: two halves compare blood cells. In mammals, red cells have no nucleus and platelets are small fragments. In birds, reptiles, amphibians and fish, oval red cells each have a visible nucleus and small nucleated thrombocytes replace platelets.

**Hemostasis**, stopping blood loss, runs in overlapping steps in mammals:

1. **Vascular spasm**: the cut vessel’s smooth muscle contracts. Narrowing steeply raises the resistance of the path out (the 1/r4 lever, roughly: a wound is no ideal tube).
2. **Platelet plug**: platelets stick to the damaged wall and recruit more platelets.
3. **Coagulation**: tissue factor, a protein exposed at the injury, starts a chain of enzymes. It ends with the enzyme thrombin converting soluble fibrinogen into insoluble **fibrin** strands. A cascade amplifies a small start into a large output, the logic of [signaling circuits](https://stemhop.com/lessons/signaling-circuits/).
4. **Clot contraction**: in human plasma clots in vitro (outside the body), platelet extensions pull on fibrin fibres and compact the clot.
5. **Fibrinolysis**: the enzyme plasmin later breaks the fibrin down.

The horseshoe crab, Limulus, has an open circulation and a clot built another way. Its blood, called hemolymph, is rich in amebocytes, cells that store clotting factors in granules. Levin and Bang (1964) reported that bacterial endotoxin makes amebocyte material clot; later work traced a cascade that ends in a gel protein, coagulin, not fibrin.

### Design problem

Do one tissue’s capillaries reabsorb fluid at their venous ends in steady state, or filter along their length while lymph returns it? Design the measurements.

> **One way to do it.** In single microvessels (Landis’s occlusion method, or perfused vessels as in Adamson and colleagues’ rat mesentery), measure filtration in the first seconds after a pressure step and again after many minutes. The revised model predicts that absorption after a drop fades as protein beneath the glycocalyx evens out again; the classic model predicts it lasts. Measure oncotic pressure as close to the outside of the endothelium as you can, not only in bulk tissue fluid. Blocking the lymphatics is weaker: swelling shows net filtration, but the classic model also leaves a small filtrate for lymph, so swelling cannot tell the models apart. Repeat in another tissue; one bed in one species does not speak for all.

## Going deeper

### The full Starling relation

In full, Jv = K × [(Pc − Pi) − σ(πc − πsub)]: Jv is the fluid flux, K the hydraulic conductance of the wall (how easily water passes; the inverse of its resistance), and σ is the **reflection coefficient** for protein: 1 if protein cannot cross at all, 0 if it crosses freely. The worked problems above set σ = 1, and the classic model also sets πsub = πi.

### Sources for the revised view

The reviews are Levick and Michel (2010) and Woodcock and Michel (2021): sustained venous-end reabsorption lacks reproducible experimental support as a general account.

### Blood numbers and clotting factors

In adult humans red cells make up roughly 36 to 54 percent of blood volume, a hematocrit that is lower in women than in men. In the platelet plug, von Willebrand factor helps platelets stick to the damaged wall, and the platelets release ADP, which recruits more platelets. Much of the coagulation chain of enzymes and cofactors is assembled on the surface of activated platelets; it ends when prothrombinase converts prothrombin to thrombin, and Factor XIII stabilizes the fibrin mesh.

---

From STEMhop (https://stemhop.com), free lessons by Christopher Deppmann. Lesson text © the authors. Practice questions: https://stemhop.com/lessons/blood-and-vessels/
