The Fluid Mosaic Model of Membrane Structure

Spec Mapping: This lesson is mapped to OCR H420 Module 2.1.5 — Biological membranes (refer to the official OCR H420 specification document for exact wording). It covers the roles of cell membranes, the fluid mosaic model and the function of each component (phospholipids, cholesterol, intrinsic and extrinsic proteins, glycoproteins, glycolipids).

Every living cell is bounded by a plasma (cell surface) membrane, and eukaryotic cells contain many internal membranes surrounding organelles such as mitochondria, chloroplasts, the nucleus, endoplasmic reticulum and Golgi apparatus. These membranes are not inert barriers — they are highly organised, dynamic structures that control what enters and leaves compartments, host enzymes, and carry out cell signalling. This lesson develops the OCR H420 Module 2.1.5 content on membrane roles and the fluid mosaic model of membrane structure and the roles of its components.

A firm grasp of membrane architecture underpins every subsequent lesson in this course: diffusion, osmosis, active transport, signalling and even how cells coordinate mitosis depend on what the membrane is and does.

1. Roles of Cell Membranes

Biological membranes perform a remarkable range of functions:

Compartmentalisation — separating the cell from its surroundings and creating specialised internal environments inside organelles.
Selective permeability — allowing some substances to cross while excluding others.
Cell signalling — receiving chemical signals (hormones, neurotransmitters) via receptor proteins.
Cell recognition — using glycoproteins and glycolipids as identification markers (the ABO blood group antigens are a classic example).
Site of biochemical reactions — the inner mitochondrial membrane houses the electron transport chain; thylakoid membranes carry photosystems.
Cell–cell adhesion — tight junctions, desmosomes and plasmodesmata in plants.
Maintenance of cell shape — together with the cytoskeleton.

Key Definition — Partially Permeable Membrane: A membrane that allows certain small, uncharged or lipid-soluble molecules to pass freely while restricting the passage of larger, charged or hydrophilic molecules.

2. Historical Context — From Davson–Danielli to Singer–Nicolson

Our current picture of membrane structure — the fluid mosaic model — was proposed by S. Jonathan Singer and Garth Nicolson in 1972. It replaced earlier models, most notably the Davson–Danielli "sandwich" model (1935), which placed protein layers on the outside of a lipid bilayer like bread around a filling.

Why was the Davson–Danielli model rejected?

Freeze-fracture electron microscopy split membranes through the middle of the bilayer and revealed proteins embedded in and passing through the lipid, not sitting only on the surface.
Flourescent labelling experiments by Frye and Edidin (1970) showed that membrane proteins move laterally within the bilayer — membranes are fluid, not static.
Different membranes have very different protein:lipid ratios, which a uniform sandwich could not explain.

The fluid mosaic model describes the membrane as:

Fluid — phospholipids and most proteins can move laterally within the plane of the bilayer.
Mosaic — proteins of varied size, shape and function are scattered throughout the bilayer, like tiles in a mosaic.

graph TD
  A[Cell Membrane - Singer-Nicolson 1972] --> B[Phospholipid Bilayer]
  A --> C[Proteins - intrinsic and extrinsic]
  A --> D[Cholesterol]
  A --> E[Glycoproteins and Glycolipids]
  B --> B1["Hydrophilic phosphate heads<br/>face aqueous environment"]
  B --> B2["Hydrophobic fatty acid tails<br/>face each other"]

3. The Phospholipid Bilayer

The structural backbone of every biological membrane is a phospholipid bilayer — two layers of phospholipid molecules arranged tail-to-tail.

3.1 Phospholipid Structure

A phospholipid consists of:

A glycerol backbone.
Two fatty acid chains esterified to the first and second carbons (the hydrophobic "tails").
A phosphate group esterified to the third carbon, often linked to a small polar molecule such as choline (forming phosphatidylcholine). This is the hydrophilic "head".

Because the molecule has both a hydrophilic head and hydrophobic tails, it is amphipathic (or amphiphilic).

3.2 Why a Bilayer?

In aqueous surroundings, phospholipids spontaneously arrange themselves so that:

Hydrophilic heads face outward, interacting with water inside and outside the cell.
Hydrophobic tails face inward, away from water, held together by hydrophobic interactions and van der Waals forces.

This self-assembly is driven by the thermodynamics of water exclusion (the hydrophobic effect) and produces a stable, 7–10 nm thick bilayer.

Region	Molecule type	Character
Head	Phosphate + glycerol + polar group	Hydrophilic
Tails	Two fatty acid chains	Hydrophobic

Exam Tip: Always describe phospholipids as amphipathic or having both hydrophilic and hydrophobic regions — not simply as "polar".

3.3 Consequences for Permeability

The hydrophobic core of the bilayer is the fundamental reason that:

Small non-polar molecules (O₂, CO₂, N₂, steroid hormones) can dissolve in and cross the membrane freely.
Small polar uncharged molecules (H₂O, urea) cross slowly by simple diffusion — though water also uses protein channels called aquaporins.
Ions (Na⁺, K⁺, Ca²⁺, Cl⁻) and large polar molecules (glucose, amino acids) cannot cross unaided — they require transport proteins.

4. Membrane Proteins

Proteins make up 25–75% of the mass of a membrane depending on its role (the inner mitochondrial membrane is protein-rich; the myelin sheath is lipid-rich). They are classified by how tightly they are held in the membrane.

4.1 Intrinsic (Integral) Proteins

Intrinsic proteins are embedded within the phospholipid bilayer. Their surfaces that contact the hydrophobic tails have hydrophobic amino acid R-groups (e.g. valine, leucine, phenylalanine), while regions exposed to the aqueous environment are hydrophilic.

Transmembrane proteins span the entire bilayer. These include:

Channel proteins — form hydrophilic pores for specific ions or molecules.
Carrier proteins — bind a specific substrate and change shape to move it across.
Pumps — carrier proteins that use ATP to move substances against a concentration gradient (e.g. the Na⁺/K⁺ ATPase).
Receptor proteins — bind signalling molecules on the outside and transmit a signal inside.
Enzymes — e.g. ATP synthase in the inner mitochondrial membrane.

4.2 Extrinsic (Peripheral) Proteins

Extrinsic proteins sit on the surface of the membrane (inner or outer leaflet) and associate with it via weak non-covalent interactions with phospholipid heads or intrinsic proteins. Many act as part of signalling cascades or the cytoskeletal attachment network.

Feature	Intrinsic (integral)	Extrinsic (peripheral)
Location	Embedded in bilayer	On surface of bilayer
Amino acid character	Hydrophobic regions in core	Hydrophilic throughout
Removal	Detergents that disrupt bilayer	Mild salt or pH change
Typical roles	Channels, carriers, receptors, enzymes	Signalling, cytoskeletal attachment

5. Cholesterol — The Membrane Modulator

Cholesterol is a small, lipid-based molecule (a steroid with four fused rings) found in most animal cell membranes. It sits between phospholipids, with its polar hydroxyl group near the phosphate heads and its hydrophobic rings nestled against the fatty acid tails.

Functions of cholesterol:

Reduces fluidity at high temperatures by restricting phospholipid movement.
Prevents the membrane becoming too rigid at low temperatures by stopping the fatty acid tails packing too closely.
Reduces permeability to small water-soluble ions and molecules by filling gaps between phospholipids.
Increases mechanical stability — mammalian membranes without cholesterol are fragile.

Plants do not contain cholesterol but have related molecules called phytosterols (such as stigmasterol). Prokaryotes generally lack sterols.

Exam Tip: Cholesterol is often described as a "fluidity buffer". It decreases fluidity when it would otherwise be too high, and increases it (by preventing crystallisation) when it would otherwise be too low.

6. Glycoproteins and Glycolipids

Some membrane lipids and proteins carry short carbohydrate (oligosaccharide) chains. These chains extend from the outer (extracellular) surface of the plasma membrane only.

Glycoproteins — protein with covalently attached carbohydrate.
Glycolipids — lipid (usually a phospholipid) with covalently attached carbohydrate.

Together they form the glycocalyx, a carbohydrate-rich "coat" on the outside of the cell.

Functions include:

Cell recognition and identification — the ABO blood group antigens are glycoproteins/glycolipids on red blood cell membranes.
Cell adhesion — binding cells together to form tissues.
Receptors for signalling molecules (e.g. hormones and neurotransmitters).
Receptors for infection — viruses and toxins often recognise specific cell-surface glycoproteins (e.g. influenza binds sialic acid).
Stabilising the membrane by forming hydrogen bonds with water molecules around the cell.

7. Fluidity of the Membrane

The "fluid" part of the fluid mosaic model is real and important. Phospholipids in each leaflet diffuse laterally roughly 10⁷ times per second, while flip-flop (transverse) movement across the bilayer is extremely rare because the polar heads cannot easily pass through the hydrophobic core.

Factors affecting fluidity:

Factor	Effect on fluidity
Increasing temperature	Increases — more kinetic energy, tails move more
Unsaturated fatty acids (double bonds, "kinks")	Increase — kinks prevent tight packing
Saturated fatty acids (straight)	Decrease — pack tightly
Cholesterol at high T	Decrease — restrains movement
Cholesterol at low T	Increase — prevents crystallisation
Longer fatty acid tails	Decrease — more van der Waals contact

Organisms in cold environments (e.g. Arctic fish, cold-tolerant plants) have membranes with a higher proportion of unsaturated fatty acids to keep the membrane fluid at low temperature — a phenomenon called homeoviscous adaptation.

8. Common Exam Mistakes

Describing the membrane as a "sandwich" of proteins — this is the outdated Davson–Danielli model, not the fluid mosaic model.
Saying carbohydrate chains are on both sides of the plasma membrane. They are on the outer (extracellular) surface only.
Calling every transmembrane protein a "channel". Channels form pores; carriers change shape; pumps use ATP — all are different.
Forgetting that cholesterol is an animal component — plant cells contain phytosterols instead.
Describing phospholipids as "polar" — they are amphipathic, with both polar and non-polar regions.
Claiming that ions can pass through the phospholipid bilayer directly. They cannot — they need channel or carrier proteins.
Writing that membranes are "fully permeable" or "impermeable". They are partially (selectively) permeable.

9. Exam-Style Questions

Describe the roles of cholesterol in the plasma membrane of an animal cell. (3)
Explain why a phospholipid forms a bilayer in an aqueous environment. (4)
Distinguish between intrinsic and extrinsic membrane proteins. (3)
Explain why the fluid mosaic model replaced the Davson–Danielli model. (4)

Model answer for (2): "Phospholipids are amphipathic: they have hydrophilic phosphate heads and hydrophobic fatty acid tails. In an aqueous environment the hydrophilic heads interact with water while the hydrophobic tails are excluded from water. The most thermodynamically stable arrangement is a bilayer, with heads facing outward into the aqueous environments on both sides and tails shielded from water in the interior. This arrangement is stabilised by hydrophobic interactions and van der Waals forces between tails."

Summary

The fluid mosaic model (Singer and Nicolson, 1972) describes membranes as a fluid phospholipid bilayer studded with a mosaic of proteins.
Phospholipids form the bilayer; they are amphipathic with hydrophilic heads and hydrophobic tails.
Intrinsic proteins span or are embedded in the bilayer (channels, carriers, pumps, receptors, enzymes); extrinsic proteins sit on the surface.
Cholesterol is a fluidity buffer — it restricts movement at high temperatures and prevents rigidity at low temperatures.
Glycoproteins and glycolipids on the outer surface form the glycocalyx and act in cell recognition, adhesion and signalling.
The membrane is partially permeable: small non-polar molecules cross freely, while ions and large polar molecules need transport proteins.
Fluidity depends on temperature, fatty acid saturation, chain length and cholesterol content.

10. Historical Development — A Paradigm Shift

A-Level depth requires you to recognise the fluid mosaic model as the current consensus and to know what it replaced. The pre-history of membrane structure is studded with paradigms that were each, in their day, taken as established:

10.1 Overton (1899) — Lipid Solubility and Permeability

Charles Ernest Overton observed that substances which dissolved readily in lipids (such as ether, alcohols, anaesthetics) entered cells far faster than substances which did not. He proposed (in paraphrase) that the cell boundary must contain a lipid layer that acts as the rate-limiting filter for molecular entry. This was the first quantitative argument that a cell membrane was lipid in nature.

10.2 Gorter and Grendel (1925) — The Bilayer Hypothesis

Evert Gorter and François Grendel extracted the lipids from red blood cell (erythrocyte) ghosts, spread them on a water surface as a monolayer, and measured the area of the monolayer. They found that the area of the lipid monolayer was approximately twice the surface area of the erythrocytes from which it had come. They proposed, in paraphrase, that the cell membrane must therefore be a lipid bilayer. (Later analysis showed their extraction efficiency was incomplete and their cell-area measurements slightly off — but the two errors fortuitously cancelled, and the bilayer conclusion turned out to be correct.)

10.3 Davson and Danielli (1935) — The Sandwich Model (Paradigm-Superseded)

Hugh Davson and James Danielli added to the bilayer the observation that membranes are not pure lipid — they show measurable surface tension that is too low for a bare lipid surface and protein is detectable on the surface. They proposed a "sandwich" structure: a phospholipid bilayer coated on both faces by a layer of globular protein. This model dominated mid-twentieth-century cell biology and was reproduced in textbooks for decades. It was later paradigm-superseded by the fluid mosaic model.

10.4 Robertson (1959) — The Unit Membrane (Paradigm-Superseded)

J. David Robertson developed an electron-microscopy refinement of Davson–Danielli — the "unit membrane" — in which all biological membranes shared the same three-layered (dark-light-dark) appearance ~7.5 nm thick. The unit-membrane hypothesis crystallised the sandwich model and pushed it for a generation. It too was superseded by fluid-mosaic thinking once freeze-fracture electron microscopy and biophysical measurements showed proteins inside, not coating, the bilayer.

10.5 Frye and Edidin (1970) — Membranes Are Fluid

L. David Frye and Michael Edidin performed a beautiful and decisive experiment. They fused a mouse cell (whose membrane proteins they labelled with a green fluorescent antibody) with a human cell (whose proteins they labelled red) to produce a heterokaryon — a single cell with two nuclei and a mixed surface. Within ~40 minutes the green and red labels had intermixed homogeneously. Proteins were therefore mobile within the plane of the bilayer; the bilayer was a two-dimensional fluid, not a static structure.

flowchart LR
  A["Mouse cell<br/>green-labelled proteins"] --> C["Cell fusion<br/>(Sendai virus)"]
  B["Human cell<br/>red-labelled proteins"] --> C
  C --> D["Heterokaryon: green and red<br/>initially on opposite halves"]
  D --> E["~40 min at 37 degrees C"]
  E --> F["Proteins fully intermixed<br/>=> membrane is fluid (lateral diffusion)"]

10.6 Singer and Nicolson (1972) — The Fluid Mosaic Model

Synthesising lipid-bilayer evidence (Gorter–Grendel), protein-embedded freeze-fracture evidence, and Frye–Edidin's protein mobility, S. Jonathan Singer and Garth Nicolson proposed the fluid mosaic model in 1972. Proteins are embedded in a fluid bilayer like icebergs in a sea, not coating it like bread on butter. This is the consensus paradigm today.

A-Level Depth: Notice the shape of the paradigm shift. Each model (Overton lipid layer, Gorter–Grendel bilayer, Davson–Danielli sandwich, Robertson unit membrane, Singer–Nicolson mosaic) was rationally defended on the evidence available at the time. New experimental techniques (freeze-fracture EM, fluorescent labelling, biophysical measurement) produced anomalies that the old paradigms could not absorb, and the field shifted. This is the classical pattern of scientific progress.

11. The Bilayer as an SVG

12. KaTeX — Permeability and the Bilayer

The partition coefficient $K$ of a solute between lipid and water, and its diffusion coefficient $D$ in the bilayer, determine the permeability $P$ of the membrane to that solute:

$P = \frac{K D}{d}$

where $d$ is the bilayer thickness (~7 nm). This is the Overton rule expressed quantitatively: lipid-soluble solutes have large $K$ and therefore large $P$ , while ions have $K \approx 0$ and therefore $P \approx 0$ .

13. Synoptic Links

Synoptic Links — Connects to:

ocr-alevel-biology-biological-molecules / lipids-and-phospholipids (phospholipid amphipathicity; cholesterol as a steroid lipid; the hydrophobic effect that drives bilayer self-assembly).

ocr-alevel-biology-membranes-cell-division / diffusion-facilitated-diffusion (the architecture established here is the substrate for every passive-transport mechanism).

ocr-alevel-biology-membranes-cell-division / active-transport-endocytosis-exocytosis (integral pumps and carrier mechanisms exploit the bilayer barrier).

ocr-alevel-biology-membranes-cell-division / cell-signalling (glycoproteins and integral receptors are the molecular substrate of signalling).

ocr-alevel-biology-exchange-transport / gas-exchange-surfaces (alveolar membrane permeability is determined by the same bilayer principles).

ocr-alevel-biology-neuronal-hormonal / resting-and-action-potentials (ion-channel selectivity in neurones depends on the integral-protein architecture introduced here).

14. Specimen Question — modelled on the OCR H420 paper format

Question (9 marks): Describe the structure of the cell surface membrane as predicted by the fluid mosaic model. Using examples, evaluate how the experiments of Frye and Edidin (1970) and the limitations of the Davson–Danielli model contributed to the acceptance of the fluid mosaic model.

AO breakdown

Mark	AO	Awarded for
1	AO1	Phospholipid bilayer with hydrophilic heads outward, hydrophobic tails inward.
2	AO1	Intrinsic (integral) proteins spanning the bilayer; extrinsic (peripheral) proteins on the surface.
3	AO1	Cholesterol between phospholipids; glycoproteins / glycolipids on outer surface forming glycocalyx.
4	AO1	Description of fluidity (lateral diffusion of phospholipids and many proteins).
5	AO2	Frye–Edidin: fusion of mouse and human cells, fluorescent labels intermix within ~40 min — proteins are mobile.
6	AO2	Davson–Danielli sandwich placed protein on the outside of the bilayer; freeze-fracture electron microscopy showed proteins embedded in / passing through the bilayer.
7	AO2	Different membranes have different protein:lipid ratios, incompatible with a uniform protein-coated sandwich.
8	AO3	Evaluation: Frye–Edidin shows the fluid component; freeze-fracture shows the mosaic component; Singer–Nicolson synthesis integrates both.
9	AO3	Recognising the model as paradigm-superseding: Davson–Danielli is not "wrong" but was overtaken by direct structural and dynamic evidence.

AO split: AO1 = 4, AO2 = 3, AO3 = 2.

Grade C model answer (~220 words)

The cell surface membrane is made of a phospholipid bilayer. The hydrophilic heads face the water on the outside and inside of the cell, and the hydrophobic fatty acid tails face each other in the middle. Proteins are embedded in the bilayer. Some proteins go all the way through (integral / intrinsic proteins) and act as channels or carriers, while others sit on the surface (peripheral / extrinsic proteins). Cholesterol is between the phospholipids and helps with fluidity. Some proteins and lipids have carbohydrate chains attached on the outer surface — these are glycoproteins and glycolipids.

The model is called the fluid mosaic model because the phospholipids can move around in the membrane. Frye and Edidin proved this by fusing a mouse cell with a human cell. They labelled the proteins on each cell a different colour. After about 40 minutes the two colours were mixed up across the whole new cell, showing that the proteins could move. Before this, Davson and Danielli had said the membrane was like a sandwich with proteins on the outside, but freeze-fracture electron microscopy showed proteins inside the bilayer, not just on the outside. So the fluid mosaic model replaced the sandwich model.

Examiner commentary: M1 (bilayer), M2 (integral/peripheral), M3 (cholesterol/glycoprotein), M4 (fluidity), M5 (Frye–Edidin description), M6 (Davson–Danielli problem). Around 6/9. The candidate names the experiment but does not analyse the mosaic vs fluid distinction or comment on protein:lipid variability. Missing the AO3 evaluative move (paradigm-supersession framing). Solid Grade C.

Grade A* model answer (~430 words)

Singer and Nicolson's fluid mosaic model (1972) describes the plasma membrane as a two-dimensional fluid of phospholipids in which proteins are dispersed as a mosaic. The phospholipids form a bilayer ~7 nm thick: each molecule has a hydrophilic phosphate head and two hydrophobic fatty acid tails, and the amphipathic geometry drives spontaneous self-assembly in aqueous surroundings by the hydrophobic effect. Intrinsic (integral) proteins are embedded in the bilayer — many span it as transmembrane channels, carriers, pumps or receptors. Extrinsic (peripheral) proteins sit on the inner or outer face, attached by non-covalent interactions to phospholipid heads or to integral proteins. Cholesterol is interspersed among the phospholipids, with its polar hydroxyl near the heads and its rigid steroid rings against the tails; it acts as a fluidity buffer, restraining motion at high temperatures and preventing crystallisation at low temperatures. Glycoproteins and glycolipids carry covalently attached oligosaccharide chains on the outer leaflet only — collectively the glycocalyx — and mediate cell recognition, adhesion and signalling.

Two strands of evidence forced the displacement of the Davson–Danielli sandwich. First, Frye and Edidin (1970) fused mouse and human cells using Sendai virus to produce heterokaryons; the mouse and human membrane proteins were tagged with antibodies bearing different fluorophores (green and red). Initially the two labels occupied opposite hemispheres of the fused cell; within ~40 min at 37 °C they intermixed homogeneously. This demonstrated that integral proteins diffuse laterally — the membrane is a two-dimensional fluid. A sandwich model with proteins coating both faces cannot accommodate such free lateral motion. Second, freeze-fracture electron microscopy split bilayers through the hydrophobic core and revealed integral proteins embedded in the bilayer as raised particles on the fracture face, not as a continuous protein coat. Together with the observation that different membranes have very different protein:lipid ratios (myelin ~ 18% protein; inner mitochondrial membrane ~ 76% protein), the uniform Davson–Danielli sandwich became untenable.

Evaluation: Frye–Edidin establishes the fluid component (lateral protein diffusion); freeze-fracture establishes the mosaic component (proteins embedded, not coating). Singer and Nicolson's synthesis (1972) integrated both into a model that has been refined (membrane microdomains, lipid rafts, restricted diffusion by the cortical cytoskeleton) but not overturned. The Davson–Danielli model was not "wrong" in 1935 — it accommodated then-available surface-tension and electron-microscopy data — but was paradigm-superseded once the freeze-fracture and Frye–Edidin evidence accumulated. This is the canonical shape of a Kuhnian paradigm shift in molecular biology.

Examiner commentary: Full 9/9. The discriminators are: the named protein:lipid ratios (~18% / ~76%), the named fusion mechanism (Sendai virus), the lipid-raft refinement of the fluid mosaic model itself, and the AO3 paradigm-supersession framing. The "fluid component vs mosaic component" decomposition is exactly the evaluative move that earns the top band.

15. Practical Activity Group anchor

Practical Activity Group anchor: PAG 8 — Transport in and out of cells anchors all of Module 2.1.5. The fluid mosaic architecture established in this lesson is the structural basis for every subsequent practical investigation of membrane behaviour. PAG 1 (microscopy) is also relevant — the historical case for the unit membrane (Robertson) rests on classical electron-microscopy preparations.

16. Common errors and mark-loss patterns

Pedagogical observations — not fabricated statistics:

Conflating "polar" with "ionic". Phospholipid heads are polar (and zwitterionic in many cases), not ionic, and the bilayer is held together by hydrophobic interactions and van der Waals forces, not ionic bonds.
Calling every transmembrane protein a "channel". Channels form a hydrophilic pore; carriers change conformation; pumps use ATP. The vocabulary is examined.
Describing the membrane as a "sandwich". That is the Davson–Danielli (paradigm-superseded) model, not the fluid mosaic model. Examiners reward the historical literacy required to know that the sandwich was replaced.
Placing carbohydrate chains on both faces of the plasma membrane. Glycoproteins and glycolipids are on the outer (extracellular) face only — this is the defining asymmetry of the plasma membrane.
Forgetting cholesterol is animal-specific. Plant membranes contain phytosterols (stigmasterol, sitosterol); prokaryotes generally lack sterols.
Writing "ions diffuse through the bilayer". Ions cannot cross the hydrophobic core directly — they cross via channels or pumps. Saying otherwise is a reliable mark-loss pattern.
Saying "the fluid mosaic model proves Davson–Danielli wrong". The correct framing is paradigm-supersession: the sandwich was rational on its data; new experimental techniques produced anomalies it could not absorb.

17. A-Level-depth misconceptions

The errors that distinguish A from A*:

Confusing the fluid and mosaic components. Fluid refers to lateral diffusion within the plane of the bilayer; mosaic refers to the heterogeneous embedded distribution of proteins. Both are needed.
Treating flip-flop (transverse diffusion) as common. It is extremely rare in pure bilayers (~once per month per phospholipid) because the polar head must traverse the hydrophobic core. Specific flippase and floppase enzymes catalyse rare biological flip-flop.
Treating the bilayer as symmetric. The inner and outer leaflets have systematically different lipid compositions (e.g. phosphatidylserine is restricted to the inner leaflet in healthy cells; its appearance on the outer leaflet is a death signal).
Conflating cholesterol with phytosterols. Plant phytosterols are functionally analogous (fluidity buffers) but chemically distinct.
Believing the fluid mosaic model is "final". The current refinement (lipid rafts, membrane microdomains, the cortical cytoskeleton constraining lateral diffusion) modifies but does not overturn it.

18. Going further — undergraduate signposts

Alberts et al., Molecular Biology of the Cell — chapter on membrane structure and lipid raft microdomains; quantitative treatment of lateral diffusion (FRAP — fluorescence recovery after photobleaching).
Berg, Tymoczko & Stryer, Biochemistry — chapter on biological membranes; quantitative partition coefficients and the Overton rule.
Singer S.J. & Nicolson G.L. (1972) — the original paper, suitable for an Oxbridge interview's "read a primary source" exercise.
Oxbridge interview prompt: "Phospholipids form a bilayer spontaneously in water. What would happen if you placed phospholipids in a non-polar solvent such as cyclohexane?" (Expected answer: an inverted micelle, with tails outward toward the non-polar solvent and heads sequestered inward.)
Oxbridge interview prompt: "If the fluid mosaic model were 'wrong', how would you go about replacing it? What new evidence would you need?" (Expected answer: anomalies in current model — lipid rafts, restricted diffusion, asymmetric leaflets — already drive ongoing refinement.)

19. Quick Recap

The plasma membrane is a fluid bilayer of amphipathic phospholipids with embedded proteins, cholesterol, glycoproteins and glycolipids.
Singer and Nicolson (1972) synthesised the fluid mosaic model from freeze-fracture, biophysical and protein-mobility evidence — it replaced Davson and Danielli's (1935) sandwich model.
Frye and Edidin (1970) showed that integral proteins diffuse laterally — the fluid part of the model.
The mosaic component is established by varied protein:lipid ratios and freeze-fracture images showing embedded proteins.
The membrane is partially permeable: small non-polar molecules dissolve through; ions and large polar molecules require channels, carriers or pumps.
Cholesterol is a fluidity buffer (animals; phytosterols in plants).
The glycocalyx on the outer face mediates recognition, adhesion and signalling.

Reference: OCR A-Level Biology A (H420) specification 2.1.5 (refer to the official OCR H420 specification document for exact wording).

Lesson 1: The Fluid Mosaic Model of Membrane Structure