MIKE MOSS SAYS:        Phil,     Winds would be perpendicular to the isobars is not for the rotation of the earth, which imposes a virtual force called the Coriolis effect. Over a significant time and distance scale typical of "synoptic" weather systems like the major highs, lows, troughs and frontal systems we show on our maps, this effect leads to winds that deflect to the right in the northern hemisphere and to the left in the southern hemisphere, with the result that winds generally circulat in a clockwise manner around high pressure centers and vice versa for lows (in the northern hemisphere). This is an idealized situation called "geostrophic flow" in which the pressure gradient force and the coriolis force come to exactly balance one another. It is most closely realized in the real atmosphere at altitudes several thousand feet above the ground and higher, beyond the direct influence of friction with the earth's surface. Even there, strongly curved flow through sharp ridges and troughs can be "cross-isobaric" in so-called super-geostrophic and sub-geostrophic flow regions, but this is not much of an issue regarding a large scale discussion of upper air patterns, and a generalization that the flow parallels the pressure contours is quite reasonable.

For surface systems, the additional factor of friction comes into play, so that the flow becomes a balance of the pressure gradient, coriolis, and frictional influences. Over a typical land surface, this results in winds that flow at about a 30-45 degree angle across the isobars toward lower pressure compared to the idealized "parallel" scenario. This angle is often more like 10-25 degrees over large bodies of water where friction is much lower. This cross isobaric flow has important implications, since flow around a high pressure center is therefore divergent on average, leading to subsidence above the high (that is, air from above sinks to replace the outward flowing air near the center of the high) and often to fair weather, while flow around a low is convergent, leading to upward vertical motion (because air collecting near the center of the low has to go somewhere, and it can't go down because of the ground) that, given sufficient moisture, can lead to clouds and precipitation.

For a more in-depth discussion with some helpful graphics, start at this address and step through the several pages that follow:

http://ww2010.atmos.uiuc.edu/(Gh)/guides/mtr/fw/pgf.rxml