Wikipedia:Reference desk/Archives/Miscellaneous/2016 October 30

Welcome to the Wikipedia Miscellaneous Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

October 30

Medical Question

What is the name of the disease when your thyroid produces something that affects your muscle, in particular your eye muscles.--71.51.161.96 (talk) 08:25, 30 October 2016 (UTC)

Hyperthyroidism often results in muscle weakness. You may also be thinking of Graves' disease, which frequently affects the eyes. Someguy1221 (talk) 08:59, 30 October 2016 (UTC)

Loeys-Dietz syndrome, also known and Marfan syndrome type II is often associated with hypothyroidism and problems with or the lack of eye muscles. Our article on this is rather small. μηδείς (talk) 22:54, 30 October 2016 (UTC)

Also see Goitre. Akld guy (talk) 09:47, 31 October 2016 (UTC)

Exophthalmia --TammyMoet (talk) 18:45, 31 October 2016 (UTC)

Do bath/shower duos ever have the knobs on the opposite side from the nozzles and faucet ?

The design I've always seen puts everything on the same side, which is a problem if the shower water is too hot, as I then need to reach through a stream of scalding water to adjust the temperature. StuRat (talk) 15:08, 30 October 2016 (UTC)

Of course. See here for some examples.--WaltCip (talk) 16:26, 30 October 2016 (UTC)

Not quite sure what that is. It sounds to me like a bath/shower mixer [[1]]. If you're talking about the sort of thing WaltCip has just posted, then yes, you can have the controls wherever you like. I think there's an assumption that you wish to control the shower while you're in it and therefore putting the controls nearby is a popular thing to do.--Ykraps (talk) 17:06, 30 October 2016 (UTC)

I guess I should explain further. A porcelain tub which also has a shower nozzle up above, at one end, aiming into it, is almost universal here in Michigan. It can be used for a bath or a shower. StuRat (talk) 17:20, 30 October 2016 (UTC)

The length of pipe between the control and the nozzle affects how soon the water temperature changes when you change the control. A long delay would make it hard to get the temperature exactly right. (Hmm, I should've cited a reference for that. Oh well.) --76.71.5.45 (talk) 18:55, 30 October 2016 (UTC)

Unless it were an electronic control... --Jayron32 23:52, 30 October 2016 (UTC)

Yes, but that would require that both the hot and cold pipes extend to the shower nozzle. It would fix the problem of overcompensation, though, where I adjust the knobs but it doesn't have the desired effect on the shower temp, so I am unsure if it's just the delay or if I didn't turn the knob far enough. Also, connecting mains power to electronic shower controls seems patently unwise, so that leaves batteries, which would need frequent replacements. Perhaps some type of mechanical cable system (behind the walls and ceiling) could be used, instead ? StuRat (talk) 20:15, 1 November 2016 (UTC)

It is common in the UK to have a shower fitted above a bath but that doesn't sound like what you're describing. Perhaps if you had a picture? All the components can be brought separately however so you could have the hot and cold supplies taken to a thermostatic mixer valve, then the mixed water piped to a diverter valve, one outlet going to a showerhead either on the wall or ceiling, and the other going to a waterspout in the wall or mounted on the edge of the bath. The components can be wherever you want them. In the meantime, one method you could use is to turn the diverter switch to the bath outlet, turn on the water and run until the temprature is correct, then switch to the showerhead.--Ykraps (talk) 05:47, 31 October 2016 (UTC)

I do that, but alas the water temperature does not stay correct. There's an interesting effect in winter that the ice cold water in the pipes is soon flushed out, but then slowly warming water comes out for quite some time, until the pipes are warmed up and the water reaches it's max temp, but then soon it starts to cool down again, as the water heater is no longer sending water at the max temp. StuRat (talk) 20:12, 1 November 2016 (UTC)

Mmmm, sounds like you've got a problem there. Of course, I don't know what type of hot water system you have but I don't know of one where that would be normal. With a combi[[2]] or Y-plan system,[[3]] that could indicate a faulty diverter valve.--Ykraps (talk) 19:47, 3 November 2016 (UTC)

Sand timer.

What is a 1min sand timer, encased in wood, with brass ends, used for. Thanks.Mark Anthony Gibson (talk) 17:35, 30 October 2016 (UTC)

Any number of things, but one use might be to time turns in a game, such as Pictionary. One advantage of a sand timer is that even very small children can use it, and know when time is almost up by when the sand is almost gone, whereas if they can't tell time, a digital timer is useless to them until it beeps. Also, you don't need to change batteries in a sand timer. The disadvantage of not being very accurate doesn't much matter here, especially if each team or player gets the same amount of time, and that time is about a minute. StuRat (talk) 17:39, 30 October 2016 (UTC)

See also egg timer. --Jayron32 20:03, 30 October 2016 (UTC)

There is some information at Hourglass and Marine sandglass. CambridgeBayWeather, Uqaqtuq (talk), Sunasuttuq 02:05, 31 October 2016 (UTC)

What is the most distant place on mainland usa from those 4 points/places, new york city, san francisco, seattle and miami?

Where is the most distant place on mainland usa from those 4 points/places, new york city, san francisco, seattle and miami?
PS: Not talking about average distance here, the place would be X km away from one of the new york city borders, Y km away from one of the SF borders, Z km away from one of the seattle borders and W km away from one of the miami borders and the SMALLEST value of this list of values (X, Y, Z and W) would be the BIGGEST one possible.201.79.56.119 (talk) 22:45, 30 October 2016 (UTC)

Draw a great circle line from New York to San Francisco and another from Seattle to Miami, the spot you want is where the two lines cross. Roger (Dodger67) (talk) 23:01, 30 October 2016 (UTC)

The point would be on the Mexican border of Texas or maybe in Kansas near Oklahoma [4]. These are far from where the great circles intersect (the line from the point near San Fransisco to Miami shows why this isn't the way to answer the question). I made the circle from Seattle smaller than the others to prevent the map from switching to the next zoom level which would make it even harder to try to tell whether it's Kansas or Texas. As can be seen here it's too far to affect the outcome. Sagittarian Milky Way (talk) 00:34, 31 October 2016 (UTC)

The shortest distance between any two points on the surface of the earth is via great circles. It stands to reason that Dodger67's response would be correct. If the red arcs on your map are great circles, then they cross in western Nebraska near the Wyoming line. I don't think the San Fran to Miami line would figure into it. ←Baseball Bugs ^{What's up, Doc?} carrots→ 05:48, 31 October 2016 (UTC)

• No, Dodger's method isn't correct. It's based on the shortest paths between pairs of the four cities, but that's not what the question asks for. Proof: imagine the question being asked with three cities spaced roughly evenly along a straight line, and one other: say Toronto, Detroit, Chicago, and Miami. Try the same method and you'll end up with a point on the straight line; but obviously the right answer (added later: if it's inside the triangle) will be roughly halfway between Detroit and Miami. --76.71.5.45 (talk) 10:28, 31 October 2016 (UTC)

This should make it clear. The circles are only 1800 km from the cities and the part of Nebraska near Wyoming is already covered. Here the circles are 2115 km from the cities and the only mainland US left is near the midpoint of the KS-OK and TX-Mex border. Seattle's circle can grow bigger than the others but this is about maximizing the minimum. At the exact kilometer number where the non-Seattle circles meet, for one to grow to improve the km number means another must shrink and break the 3-way tie for smallest with a worse value. Therefore the maximum minimum is in Kansas or Texas. The line from Miami to near San Francisco was just a proof by exaggeration that the intersection of two great circles is not the way to find the answer. If an alternative interpretation is the smallest has to also be biggest then x, y, z, and w would have to all be equal and we've already seen the non-Seattle ones run out of room first if they grow at equal size. Sagittarian Milky Way (talk) 07:11, 31 October 2016 (UTC)

It should be somewhere on the bisector of the SF–Miami line. NY and Seattle might be too far north/east/west to figure into it much. Except that they might rule out northern Minnesota, which looks like it might be the answer if only SF and Miami were considered. So I'm going to guess somewhere in Big Bend National Park. --Trovatore (talk) 06:13, 31 October 2016 (UTC) as soon

That's where my two candidate points are (the bisector is slightly Miamiward of Big Bend actually). New York does figure into it if it's Kansas but if you adjust the kilometer constant you can make it look like it's probably Texas but the map's too small to be sure. Someone could get a better map scale in Google Earth but it's such a close call that if Google Earth uses spherical geometry the answer could be wrong. Sagittarian Milky Way (talk) 07:20, 31 October 2016 (UTC)

Wow, that is really close. But adjusting your URL to 2145 km, it looks to me like the intersection of the SF and Miami circles creeps over the Mexican border, but stops short of the NY circle. So I'm going to say somewhere near Manchester, Oklahoma, but over the border into Kansas. --Trovatore (talk) 07:27, 31 October 2016 (UTC)

Okay, best chance with this site (which uses the WGS84 ellipsoid like GPS). I opened Google Maps satellite, clicked as close to the center of gravity of the skyscraper district as I could and pasted the coordinates here (the previous maps pasted what Google Search pops up for [city] latitude longitude which aren't actually bullseyes for the skyscraper district. i.e. "Midtown Manhattan latitude longitude" gives 6th Avenue and 42nd Street which is .2 miles from where I put the fuzzy center of Midtown (it's obviously a good idea to try that before "New York City latitude longitude" because Midtown's more of a downtown than Downtown even though it's uptown of Downtown but they might give the coordinates of Downtown which is 4 miles downtown of downtown). I looked at that part of the Mexican border on Google Maps, imagined a best-fit curve for the sharpest contrast between dark and light marking the circles and.. I think it's Texas. But it's almost too hard to tell. Sagittarian Milky Way (talk) 09:04, 31 October 2016 (UTC)

I decided to try some brute force. I already have a program that calculates an approximation of the distance between two places based on latitude and longitude (it's approximate because the way it allows for the Earth not being spherical is wrong), and I ran it on positions every 10' from 33° to 44°N and from 90° to 106°W. I then reran it on the best area, checking positions every 1'. I also checked positions near the Northwest Angle of Minnesota and on the Texas-Mexico border. (Based on Wikipedia, I took the position of the four cities as 40°43'N 74°0'W, 37°47'N 122°25'W, 47°37'N 122°21'W, and 25°47'N 80°13'W respectively.)

With the distance computed to the nearest kilometer, it could not distinguish between five "best" solutions, grouped around 37°13'N 98°16'W, which is a couple of miles southwest of Attica, Kansas—more or less confirming the results by others above. The "Measure distance" function in Google Maps shows that position as 2,126.67 km from (the coordinate position I used in the last paragraph for) New York, 2,125.45 km from San Francisco, and 2,125.43 km from Miami, so I think it's a pretty good answer, especially when you consider that a city is larger than a single point. (It's farther from Seattle at 2,276.38 km; generally you would expect three city distances to be equal and one to be farther if the position is within the quadrilateral formed by the three cities.) --76.71.5.45 (talk) 10:54, 31 October 2016 (UTC)

Wikipedia gives coordinates for a part of Miami in the middle of nowhere. San Francisco's coordinates are neither in the geographical center of the city or the central business district. It gives the coordinates of City Hall for New York. People need to stop giving the coordinates of City Hall for NYC. It hasn't been the geographical center in 2 centuries, it's not in the middle of the skyscrapers, and even if it was those skyscrapers are all the way at the tip of the island when the center of gravity of non-poor Manhattanites is much further uptown and there's a bigger central business district 4 miles from it (center-to-center). If we simplify the non-poor part to "96th Street down" the tip is "negative 41st Street" and the center is 27-28 Street in greater Midtown. Manhattan averages about half the normal width till 1st Street so you get a center of non-poor Manhattan closer to 38th Street which is between the Empire and Chrysler Building. The Ivy League Columbia University takes up the West 110s and the stuff in between is gentrifying so maybe push that even closer to central Midtown. And voila!, what a fitting name, Midtown's in the middle of town. Sagittarian Milky Way (talk) 12:58, 31 October 2016 (UTC)

@Sagittarian Milky Way: Kilometre zero#United States and Wikipedia:Obtaining geographic coordinates might be more useful places to progress this discussion. jnestorius^(talk) 15:48, 3 November 2016 (UTC)

That the great circles approach is wrong at least in general is easy to see when considering the point that is farthest from Boston, New Haven, Providence and Philadelphia. Great circles indeed give us the minimum distance between points, but the OP is interested in the maximum. --Stephan Schulz (talk) 08:40, 31 October 2016 (UTC)

• Would it be just as simple as drawing the quadrilateral whose corners are the 4 cities, then finding the centroid of that quadrilateral? I know that we'd have to take into account the fact that it isn't a true quadrilateral, but it still seems pretty straightforward. --Jayron32 10:45, 31 October 2016 (UTC)

• No. Consider the example I cited above, where three cities are in a straight line. If Detroit was exactly halfway between Toronto and Chicago, then the centroid would be 1/3 of the way from Detroit to Miami, but as I said, the solution point would be more like halfway between them (added: if it's inside the triangle). --76.71.5.45 (talk) 10:54, 31 October 2016 (UTC)

• Also consider my example, where the point in question clearly lies outside the area of the tetragon described by the 4 points. Indeed, for a mathematically simpler degenerate example, look for the point in the continental US that is farthest from Boston, Boston, Boston and Boston ;-). --Stephan Schulz (talk) 11:07, 31 October 2016 (UTC)

• OK Then. Could you draw a circle of radius "X" centered on each city, where "X" is the distance from that city to the point in the continental US farthest from that city, then find the point that is the shortest distance from those 4 circles? --Jayron32 12:55, 31 October 2016 (UTC)

Hypothetically, yes. But you'd need an enormous compass, as it could only be done in 3-dimensional reality, not on a conventional 2-dimensional map. That's if you wanted precision; and from the above answers it looks like pin-point precision is important in this question. Maybe a large globe might be suitable. You'd still need an outsize compass. -- Jack of Oz ^{[pleasantries]} 18:03, 31 October 2016 (UTC)

How do you measure distance from a big city with pinpoint precision? From the nearest point on the city's legal boundary? From an arbitrary density contour? —Tamfang (talk) 20:48, 31 October 2016 (UTC)

It's a question of definition. No matter what methodology is used, before this exercise is attempted the measurement point must be agreed. With cities of this size, it makes a difference whether you measure from the legal boundary or some central point in the CBD, for example. -- Jack of Oz ^{[pleasantries]} 21:19, 31 October 2016 (UTC)

By the most literal reading of the original question, we're supposed to maximize the minimum distance to any point in any of the four cities. So for example Saggitarian's "circle" centered at New York should really be the union of all those circles I should actually say "disks"; a "circle" is a curve, not a region centered at any point in the Five Boroughs. --Trovatore (talk) 21:42, 31 October 2016 (UTC)

(My guess, by the way, is that that will make the answer much clearer — the southern tip of Staten Island should eliminate the Oklahoma/Kansas point from consideration.) --Trovatore (talk) 21:47, 31 October 2016 (UTC)

• The point sought will be on the Voronoi diagram defined by the four cities: either at one of its vertices or where an edge crosses the border. In general there's probably no way to avoid examining each such point separately; but for four cities there should be only four vertices and six edges. —Tamfang (talk) 20:57, 31 October 2016 (UTC)

• Forget great circles. To clarify, look at the problem of the farthest point in the mainland US from The following four cities: Manhattan, Brooklyn, Queens, and Albany, all in New York state. the farthest point is probably near San Diego. This means that the Voronoi diagram is probably the correct approach. -Arch dude (talk) 03:50, 2 November 2016 (UTC)