Phospholipids will spontaneously associate into an extended monolayer at the interface between a saline solution and oil. They will behave similarly at the air surface of any aqueous solution, where the hydrocarbon tails form a very thin oil film and the polar heads project into the solution. Given this behavior it's not too hard to imagine how a stable phospholipid bilayer might be formed within an aqueous solution from two monolayers laid "back-to-back". The hydrocarbon tails from each monolayer would touch, and together these tails would constitute a hydrophobic core for the bilayer. The polar heads of each monolayer would interact with each other and with the adjacent solution; these interactions would exclude the tails and help stabilize the core.
Real bilayers can be constructed from aqueous suspensions of many amphipathic lipids in high concentration. If the suspension is rapidly aggitated, microscopic spheres called liposomes are formed. These liposomes are vesicles consisting of a continuous lipid bilayers that effectively separate the saline solution into interior compartments and exterior solution. In fact, liposomes look like very simple, minature cells. But are plasma membranes surrounding real cells really bilayers?
A good scientific way of answering this question would be to make a specific, testable prediction based on current information and then to test the accuracy of the prediction. So what do you know so far about plasma membranes? Yes, they contain lots of amphipathic lipid (Chaps. 1 and 2)! And what do you know about amphipathic lipids? They'll assemble in aqueous environments into stable, organized structures such as monolayers and liposomes (Chaps 3 and 4). Now, make a testable prediction!
Gorter and Grendal, the "founding fathers" of our current model of plasma membrane structure, made one such prediction. They reasoned quite simply that if a plasma membrane were really a bilayer then its surface area should be half that occupied by all its amphipathic lipids spread out in a monolayer. They then made independent measurements of the surface areas of different mammalian erythrocytes. They also extracted the lipids from erythrocyte membranes, spread the lipids out at an air/saline interface and measured the surface areas of the resulting monolayers. Comparison of these two measurements indicated a cell:monolayer surface area ratio of about 1:2, confirming for Gorter and Grendal the bilayer nature of the plasma membrane.
But where's all the membrane protein? Continue on to the [next chapter - under construction] to examine this question and to explore the nature of integral and peripheral proteins. [If you've been following this thread from a LabBook presentation of Membrane Fluidity, you can easily return to this topic.]