I had an email exchange with Marty Hardison a few days ago. He brought up an argument in favor of short top-bars, which hadn't occurred to me before. Nor had I seen it in the discussions here.
Here it is. Bees like to build curved comb--the comb is not typically a perfect plane, but is somewhat cupped. The longer the bar, the greater the total divergence from a a plane, and therefore the more trouble (for the beekeeper) with trying to fix combs which don't stay tidily on their own bar.
But, with a shorter bar, you're getting a shorter section (from end to end) of the natural curve, and therefore it coincides more closedly with the straight line desired (for ease of management) by the beekeeper.
(Think of the tangent of an arc. Draw perpendiculars from the two ends of the tangent toward the arc. These perpendiculars increase in length as the length of the tangent increases. The tangent is the bar, the arc is the comb. This is the mental picture that makes Hardison's argument convincing to me.)
He's passionate about this, because he thinks a lot of beginners who start with long bars get frustrated and quit on account of comb management troubles.
He also would agree that really long bars are more prone to comb failure, and says that the longer the bar, the more severe the taper of the hive needs to be, to mitigate stresses on the comb. With a short bar, you can have hive sides which are nearly vertical, and with a long bar, you'll need more sloped sides. In light of this, I'm not sure what to make of the Tanzanian hives with long bars, and vertical sides, and operating in hot weather. Maybe it comes back to the issue of entrance size--a big entrance on such a hive would be a disaster, because the bees couldn't control the airflow.
[This message has been edited by dmcdonald (edited March 08, 2004).]