Why the horse-shaped house is a problem patent

In a recent post, I mentioned a patent on a horse-shaped house (U.S. patent 5,564,239). Here is a very thorough analysis from a reader: The patent for a horse shaped sculpture may also be a good example of a weak patent examination process in action. The first cited patent - from 1882 - really sets forth the entire concept of an animal shaped building. Including stating "The building may be of the form of any other animal than an elephant, as that of a fish, fowl, & etc." That description pretty clearly includes a horse, and anyone looking at the patent - or the actual buildings that resulted, would surely realize that the building could be shaped like a horse, or a cow, or any other animal. Isn't it obvious to surround a building (of any sort) with a garden and/or a moat, and/or a fence? Lets see - gardens have surrounded buildings for centuries, and we've all seen castles from 500 years ago which are surrounded by water (and even have gardens in some cases - and even bridges over the moats). Is it novel to make the outline of the garden, or the moat, octagonal? Wouldn't that be an obvious extension, of say, a pentagonal or hexagonal shape? In fact if I wanted to spend an hour or so I bet I can find an octagonal moat surrounding a building. And thus with the prior art you have the '239 patent. But it is an amusing patent. I remember visiting Lucy the Elephant - built from Lafferty's 1882 design in Margate NJ. It was a decrepit relic when I was young, but has since been restored. See the story at: http://www.lucytheelephant.org/ The web site includes historical descriptions of other similar structures, how they were used for amusement purposes, etc. By the way -- Lucy the Elephant is surrounded by a rectangular fence, with an opening. Is it really an invention to make the fence octagonal instead of rectangular, rhetorically speaking?