Along with Andrew and Matt's comments, I have two others: 1) When I was doing a lot of survey-type visual observing---always with the same telescope and same magnifications as per Steve, I found that after going after all too many objects that were really too faint for convenient viewing, I would finally come across an "ordinary" galaxy bright enough to be in the Shapley-Ames catalogue, and would describe it as "large", even though it was only 2' across. It was simply that it was so much larger than the 20"-30" stinkers I'd been looking at previously. Fortunately I have always tried to give size estimates in angular units, tied to nearby field stars if possible, so that the subjective side could be controlled with real numbers. 2) I think if you go through the original GC (not the NGC) and make a list of a representative sample of object size-categories assigned by Herschel, then match those with reliable angular diameters (say using galaxies or globulars, where consistent sizes are well in hand), I think you'll find that "large" ranges over a factor of at least five if not ten in angular size. I recall there being an S&T article (maybe Scotty Houston column?) showing exactly this. Thus I think the argument that "I use roughly the same magnification as Herschel" produces consistent results is probably spurious. The test, naturally, would be to take Steve's observations and apply the same method (what range of actual sizes does "large" apply to?) to look for the consistency. \Brian -- See message header for info on list archives or unsubscribing, and please send personal replies to the author, not the list.