杏吧原创

A pinch of this, a squirt of that. David Tompsett was a craftsman who had
mastered a tricky process using methods of his own devising. His unique skill
was in creating solid casts of delicate hollow structures: blood vessels and
airways, animal or human. The results of his work, mostly from the 1950s, are on
display at the Hunterian Museum in London and are as good to look at as they are
to learn from.

While Tompsett was busy filling in the tubes of dead creatures, Colin Caro
was beginning to discover how fluids flow through them while they鈥檙e alive. A
physician turned physicist, Caro has since spent years studying the fluid
mechanics of the blood system. Lately he鈥檚 become interested in how gases move
through the airways, and he had a hunch that sent him to the Hunterian Museum to
examine Tompsett鈥檚 castings. Physiology in life was about to be illuminated by
anatomy in death . . .

One moment you have a filigree of fine, soft pipework, once part of a
living, breathing human being. The next, you have a perfect replica, a network
of delicate tubes that is identical in all respects but one: it鈥檚 solid, fixed
for posterity. Eat your heart out, Damien Hirst.

The business of making casts of the body鈥檚 finest plumbing systems, by
injecting them with liquids that then solidify, took a long time to perfect. The
surgeon John Hunter tried it in the 18th century using curdled milk, gelatin and
even molten metal鈥攚hich didn鈥檛 do a lot for the specimens. The 1930s saw
the invention of liquid resins that could be hardened using chemical catalysts.
These offered new possibilities. One man, David Hughes Tompsett, brought them to
perfection.

Tompsett spent four decades at the Hunterian Museum of the Royal College of
Surgeons, acting for most of that time as prosector鈥攑erforming anatomical
dissections for the benefit of medical students. So entranced was he by the
possibilities of resin casting that towards the end of his career it became his
main preoccupation.

In principle the technique of Tompsett鈥檚 鈥渃orrosion casting鈥� is simple. Mix
up the liquid resin according to his instructions and inject it into the space
you wish to cast. Leave it to harden, then dissolve the surrounding tissue by
dunking the whole thing in a bath of concentrated hydrochloric acid. What
remains is a fragile 3D model of the layout of the blood vessels or the airways.
Needless to say, in practice, sticking to Tompsett鈥檚 formulas no more guarantees
success than following recipes in the latest celebrity cookbook ensures a
perfect meal.

The less viscous the resin, the further it penetrates, flooding into even the
tiniest passages and spaces. Make it too thin, however, and it will seep through
everything and harden to a useless blob. When casting the lungs, the trick is to
give the resin enough time to reach the narrowest airways before it begins to
stiffen鈥攂ut not so much that it oozes into the alveoli, or air sacs, that
make up most of the volume of the lungs. Tompsett perfected the art, but when he
died 10 years ago, so did much of the art鈥� although the models remain.

Switch from Tompsett鈥檚 back room at the Hunterian Museum to the glass and
steel of the Bagrit Centre at Imperial College, and you鈥檒l find Colin Caro. It鈥檚
here that Caro has proved that blood vessels are far more than just tubes. Back
in the 1960s, he and his colleagues showed how the fatty deposits that clog up
our arteries accumulate where the blood flow is most sluggish. Think of the way
that sand and mud build up on the inside banks of curving rivers and you can see
why the geometry of the blood vessels has the potential to create problems.

Natural selection, ever a force for improvement, has helped to mitigate the
damage. As Caro points out, vessels tend to bend and curve not in two dimensions
but in three, following a corkscrew path. This helical tendency brings a
swirling and scouring motion to the flow, and increases movement in regions
where the blood would otherwise be sluggish.

But fast-moving blood also imposes shear stresses on the cells lining the
arteries. Too much shear is bad news for blood vessels. They respond by making
chemicals that influence all sorts of things, from blood coagulation to vessel
diameter. A swirling movement modifies the flow and evens out the extremes.

A few years ago, a friend of Caro鈥檚 with a research interest in respiration
invited him to give a lecture. Out of politeness Caro felt he should talk about
breathing as well as circulation, and this got him thinking about the 3D
geometry of the airways. He wondered if they too generated this swirling flow
pattern鈥攁nd, if so, how.

The airways that enter the lungs branch out like a tree: a succession of
Y-shaped junctions in which the arms become successively narrower as they
divide. Caro reckoned that the simplest way to make the air swirl as it flowed
into the lungs would be to set each of the Y-junctions at right angles to the
one before it.

Luckily, he already had a model he could use to test this hunch鈥� a
piece of kit he鈥檇 put together 30 years earlier during his work on blood flow.
It consisted of a branching network of transparent tubes thinly coated with a
paste of litmus. As schoolchildren learn in virtually their first chemistry
lesson, litmus is blue in alkaline conditions, and red in the presence of
acid.

Although the equipment was originally designed to be filled with liquid, it
worked just as well with gas, and Caro selected an acid vapour to test his
theory. As he pumped it through the tubing, the acid slowly turned the litmus
from blue to red, with the rate of change depending on the flow conditions. When
the Y-junctions were all in the same plane, the pattern of colour change
indicated a linear flow. But when successive junctions were at 90掳 to each
other, the flow became swirling.

So much for bits of tubing. What about the layout of real airways? A visit to
the Hunterian Museum and a close inspection of Tompsett鈥檚 models provided the
final evidence. The planes of successive junctions in animal and human airways
are indeed set at roughly 90掳 to each other. QED.

The biological significance of swirling flow in the airways is still not
clear. But as Caro points out, it will certainly affect the diffusion of heat
and gases across their walls, and also determine where dust particles are
deposited in the lungs. So the findings may interest drugs companies and asthma
patients, because the flow pattern of air into the lungs is a factor in
determining how much of an aerosol drug is deposited, and where. Geoff Watts