HOW WELL DOES A BIRD'S nest cut the chill of a winter storm? How strong a breeze does it take to force a leaf-cutter ant to adjust its grip on its spoils? Answers to such questions can be found if you can measure low wind speeds in tight places. Unfortunately, anemometers using hemispherical cups mounted on a vertical shaft are bulky, and many are good only for wind speeds above about five meters per second (11 miles per hour). An approach more suitable for lower wind speeds is thermal anemometry, which determines wind velocity based on the degree to which flowing air cools a heated piece of metal.

Professional thermal anemometers rely on tiny, red-hot wires and can sample air speeds one million times every second. But such systems are expensive, and the wires often break. The more economical method described here relies on two small metal balls and a few dollars worth of electronics. One of the balls is heated by an electric current flowing through a resistor. The temperature difference between the balls indicates the wind speed to within a few percent, and the device can operate in a space as small as a peanut butter jar.

Aluminum balls 1/2 inch in diameter make the best anemometers. Aluminum conducts heat much better than steel and bronze do, and it is not easily weathered like copper and brass. It can also take a high polish and is a good reflector of both visible and infrared radiation, making the anemometer insensitive to direct sunlight.

The balls will need to be protected. Polished aluminum scratches easily, and the marks will alter the thermal properties of the balls. Although a coat of white enamel will shield them, it will also insulate the balls somewhat, making them respond more sluggishly to changes in wind speed. A better solution is to plate the unpolished balls with gold, which is extremely reflective and surprisingly durable. Plating costs less than you may think; my electroplater charges $1.50 per square inch.

Prepare the balls by drilling 3/32-inch diameter holes in them; go all the way through one ball (the one that will be heated) and about 3/8 inch into the second. Electrically insulate the leads of a 100-ohm, 1/4-watt resistor (1 percent tolerance) with a coat of latex-based enamel. When the leads are dry, insert the resistor into the ball that will be heated, allowing the leads to stick out from either end. Cement the resistor in place using a dab of low-viscosity aluminized epoxy, which provides good thermal contact between the ball and the resistor.

To heat the resistor, use a 7805 integrated circuit, which provides five volts to the resistor. You can power the chip with any direct-current voltage between five and 35 volts. This circuit will deplete a nine-volt alkaline battery in about five hours, so consider using an adapter or large lantern batteries. Also connect a heat sink to the back of the chip; I used a dab of aluminized epoxy.

The temperature difference between the balls is measured by a device called a thermocouple, which consists of two wires, made from different metals, joined together. Copper and constantan wires are most commonly used; you will need about four inches worth. Strip 3/8 inch of insulation off both ends of each wire and twist them together to form two junctions. Electrically insulate the junctions by dipping them in enamel. At the center, cut the copper wire only and solder the two ends to separate copper wires. Insert one junction into each ball and seal with aluminized epoxy.

Bend a 12-inch piece of 1/4-inch-diameter copper pipe tubing into an S-shape and secure some wire across the top opening of the 5. Mount the balls a few inches apart on the stiff wire [see illustration below]. The voltage signal from the anemometer is boosted with an operational amplifier (type 741) and read with a digital voltmeter. With these components you can detect wind speeds as low as 0.1 meter per second.

To calibrate the anemometer, you will need to measure the output voltage at a number of known wind speeds. One way to do that is to compare the readings against those of a cup anemometer. This approach will be accurate down to only about five meters per second because of the insensitivity of the cup anemometer. You can extrapolate to lower speeds, but such a projection will likely be incorrect because the rate at K; which the heated ball cools changes in low winds.

A more precise means of calibration is to pass air currents of known speeds over the device-more specifically, to place the instrument on a rotating platform. The platform can be made from a motor cannibalized from an old ceiling fan. Such fans are ideal because they come with a mountable base and speed control, and they rotate at slow, safe rates, no faster than five revolutions per second (other fans spin much too quickly to be safe).

A ceiling fan's controls are, unfortunately, mounted on the wrong side for our purposes, so you will have to rewire the switches to the back of the motor and extend the leads. Also for safety, mount the motor's speed-control circuitry inside a metal project box and, for finer adjustment, to wire in a dimmer switch as well.

Mount a meter stick on the motor housing [see Figure 1]. Clamp the anemometer to one arm of the meter stick and mount all the electronics on the other. I duct-taped my voltmeter just beside the center of the motor and read it as it spun. Although easy at slow speeds, reading the voltmeter this way becomes tough at five revolutions per second. At high speeds, you will need to read the meter with a strobe light.

As the anemometer rotates, its speed (or alternatively, the velocity of the air rushing over it) is the circumference of the circular path times the frequency of the rotation, in cycles per second. That is, the speed equals 2Rf, where R is the distance from the center of the motor to the anemometer, and f is the frequency. The rotation frequency is easily measured using a bent playing card stapled to the underside of the arm. Mount a short dowel so that the card will strike it every time around; it will produce a sharp sound. Count the number of clicks over some interval. The frequency is the number divided by the measurement time. By selecting both the anemometer's position on the arm and the motor's rotation rate, I could create any wind speed from 0.1 to 22 meters per second.

Do all your calibrations in a closed room. Seal the windows and doors; do not walk about during the trials. The device is somewhat sensitive to ambient temperature, so make sure to calibrate it on both a cold morning and a hot afternoon. Thereafter, make certain to record the air temperature whenever you are in the field.

Once the anemometer is calibrated, you can explore the subtle interplay between many animals and their environments or measure air currents anywhere inside a building, cave or large machine. Amateurs who document these "microclimates" stand shoulder to shoulder with professionals. It is an exciting area, ripe for original work from all comers.

Hot-ball anemometry kits are available for $45 from the Society for Amateur Scientists, 5600 Post Road, #114-341, East Greenwich, RI 02818. This offer expires October 1, 1996. Additional construction tips are available for $2 from the above address or may be accessed free on the society's World Wide Web page (http://www.sas.org/) or in Scientific American's area on America Online.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.