Most bees have struck a pretty rich deal with the plants they frequent. Lured by the promise of a nectar feast, the insects, in turn, pollinate their flowers. But the pact between male orchid bees and their preferred blooms looks much more onesided. In return for their nuptial role, the male's only reward is a dab of the orchid's scent. But it wasn't the bee's fragranceharvesting habits that attracted Michael Dillon and Robert Dudley to the insects. It was their remarkable size range. Some orchid bees weigh in at 50 mg, while others tip the scales at 1 g. And they're all champion hoverers. Which made Dillon and Dudley suspect that the insects were ideal for studying the way that creatures scale up the mechanical forces that keep them aloft. Dillon headed south to the insect's homes in Panama, where he spent three months collecting bees and testing their weightlifting prowess to get to grips with allometric scaling in insect flight (p. 417).

Dillon explains that trapping the bees was easy, all he had to do was soak blotting paper in each species' favourite scent; the bees came flocking. The next trick was to attach a tiny beaded string to the insect's body, without getting stung. But this wasn't as tricky as it might sound. Only males are attracted to the orchid's scent, and they aren't armed with painful stings, although the odd female did turn up from time to time, making the encounters more lively.

Ready for their weightlifting task, Dillon filmed the insects as they lifted the beaded string until it became too heavy, and then began descending gently back to earth. Dillon was most intrigued by the moments before the weight became too great; he knew that was the point when the insects were exerting their maximum force. All Dillon had to do was measure the length of the string to measure the insect's maximum force.

Back in the lab in Texas, Dillon was faced with hours of hovering flight footage that needed painstaking analysis before he could begin to understand how the insect's aeronautic abilities scaled with size. But eventually his patience was rewarded. He realised that the larger insects were generating larger forces as they hovered, but when he looked at the force per unit mass, he realised that the larger insects were generating much less force per mg of muscle than their tinier cousins. So even though the larger insects had scaledup their muscle size pro rata, the power generated per unit mass by the flight muscles had declined. And when Dillon compared the insect's wing sizes, the larger bees' wings were much bigger than if they were in the same proportion to their bodies, probably to compensate for the larger bees' relatively feeble muscles.

`It was neat to see the mass-specific force dropping with the increase in muscle mass' says Dillon, and adds that `these data sit well' with the insect's antics in their forest homes. `The little ones hover on a dime' he remembers `while the big ones lumber along. You can hear them coming.'