Wikipedia:Reference desk/Archives/Science/2009 September 20

From Wikipedia, the free encyclopedia
Science desk
< September 19 << Aug | September | Oct >> September 21 >
Welcome to the Wikipedia Science 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.


September 20[edit]

Breath-Controlled Automobiles (not the toy cars)[edit]

Either my google-foo has vanished or I dreamt this up. Somewhere in a long investigation of adaptive devices for paraplegics and quadraplegics/tetraplegics who are drivers, I seem to remember reading about investigations of and even prototypes of breath-controlled cars. There certainly are breath-controlled and voice-controlled wheelchairs, but I can't find anything on automobiles. Does anyone know where I could find some information? // BL \\ (talk) 03:32, 20 September 2009 (UTC)[reply]

Not strictly relevant but Hilary Lister just sailed round Britain in a breath-controlled boat. If you google her you might find out about something relevant about her control system. Tinfoilcat (talk) 10:58, 20 September 2009 (UTC)[reply]
Suck and puff switches, or Sip-and-puff, e.g. these, are found on many quads' wheelchairs. Nimur (talk) 15:30, 20 September 2009 (UTC)[reply]
In the L'il Abner comic strip, he once drove the "Nomoto Car," invented by a "Mr. Nomoto." It had no engine, and was powered by the driver breathing in and out, with the chest motion coupled to the wheels. With fast enough huffing and puffing, Abner could go quite fast. Edison (talk) 18:53, 20 September 2009 (UTC)[reply]
This seems like an awfully dangerous idea! Sure, you could have some kind of interface that would allow steering, gas, brake and gearshifter (plus lights, windshield wipers, turn signals, etc) to be driven by some kind of breath-controlled menu-based interface - but how fast could someone with so little dexterity actually get from (say) turning on their windshield wipers to pressing on the brake, releasing the gas and steering out of the way in the event that a little kid runs out between two parked cars right in front of the vehicle? Worse still - if the person is so disabled that they cannot turn their head and body to control the car - what hope is there that they'd be able to adequately look out of the side windows or look towards the rear when joining traffic from a freeway ramp or something?
A suck/blow switch is essentially a single analog input - and because of the necessity of ceasing to blow when you run out of breath or need to swallow saliva or lick your lips - you don't even have full control of that single input. With something as slow as a powered wheelchair - you can simply have the thing stop when the breath input ceases to be applied. You could easily imagine a wheelchair that travelled slowly forwards as long as pressure is applied to the breath tube, turns left when more pressure is applied, turns right when you suck gently on the tube and reverses when you suck hard on it. As soon as you cease sucking or blowing, the wheelchair could simply stop. That kind of an interface works fine at 2 miles per hour with everyone around the user being aware that they are severely disabled and need to be given space to manouver.
But you can't have a car travelling at 70mph go into a full-scale emergency stop every time the user needs to take a breath! Sure, it is technologically possible to slowly control a car with such a minimal interface - but is it sane to attempt it? Hell no!
What is required to give such people the mobility they'd like to have is something like the cars in the DARPA Grand Challenge that can essentially drive themselves. The user would then only be required to use the suck/blow interface to make high level decisions - like "turn left at the next intersection" that could be made in plenty of time to let the robotic systems handle the details of changing speed, steering, avoiding other cars, etc. Better still - design it so the car knows where it's going and is capable of driving the entire way without any driver input whatever - then all the user has to do with the breath interface is select between entries within a list of pre-determined destinations...just like you select your destination with a navigation computer.
But those DARPA vehicles are a LONG way from being safe enough to let loose on the roads.
SteveBaker (talk) 19:21, 20 September 2009 (UTC)[reply]
I wouldn't say "a LONG way..." It's not so very uncommon around these parts. Last summer while biking over by the Horse Farm and Robot Car Ranch, I saw a team of computer science nerds in a chase-car holding an SUV-mounted HD camera and going after Stanley, who was out driving, unmanned, on a very real and very public roadway... as far as I know, it was no accident, but then again Stanley might have escaped... Nimur (talk) 20:19, 20 September 2009 (UTC) [reply]
It's hard to imagine how that could possibly be legal. Heck, you can't legally drive a car (in the USA) where the engine starts itself without a direct command from the driver! With that degree of legal luddism - I can't see fully autonomic cars being legal for an awfully long time. Nor should they be! The potential for tiny software bugs to kill people in large numbers is too high to be tolerable. There would need to be much more extensive testing and certification by some kind of federal agency before these things could be trusted on our roads. SteveBaker (talk) 21:52, 20 September 2009 (UTC)[reply]
Well, cars with human drives manage to kill people in large numbers. But yes, there should be much more extensive testing and certification [...] before these people should be trusted on our roads. ;-) --Stephan Schulz (talk) 16:58, 22 September 2009 (UTC)[reply]

instantaneous speed question[edit]

An object is moving with uniformly increasing speed. It's speed at time zero is 0m/s. If it's average speed during the first second is 14m/s at what time will it's instantaneous speed be 14 m/s?

I know the answer is 0.50 seconds, but i don't know how you get it.

Average speed = (vinitial + vfinal) / 2 for constant acceleration. Therefore vfinal = 28 m/s, and its velocity would be 14 m/s halfway through. Not an instantaneous answer, but what do you expect for free? Clarityfiend (talk) 04:32, 20 September 2009 (UTC)[reply]
v = vinitial + a*t. 14 = 0 + 28*t -> t=0.5 if you want more specifics. Clarityfiend (talk) 04:49, 20 September 2009 (UTC)[reply]

how did you get 28? isn't it supposed to be 0m/s plus 14m/s divided by 2 which equals 7?

It's (0 + vfinal)/2 = 14 m/s. vfinal = 28 m/s. Rckrone (talk) 05:59, 20 September 2009 (UTC)[reply]
same answer told a different way
Uniformly increasing speed means constant acceleration - thus v=at , the average speed = ʃat.dt /t = 0.5at
So 14 = 0.5a1 which gives a=28m/s^2
So for v=14m/s you need to solve v=at , or t = v/a =14/28 =0.5s
The main problem is knowing how to calculate the average speed which requires integration (the area under the speed/time graph) - did all of that make sense.83.100.251.196 (talk) 09:34, 20 September 2009 (UTC)[reply]

i still don't see how you got 28.

Step by Step solution
Step 1. "uniformly increasing speed" is "constant acceleration"
Step 2. V_initial = 0 m/s
Step 3. V_average = 14 m/s
Step 4. V_average = ( V_initial + V_final )/2
Step 5. Delta_t = 1 s
Step 6. Acceleration = (V_final - V_initial) / Delta_t
step 7. V(t) = V_initial + Acceleration * t
Step 8. Find t when V(t) = 14 m/s

202.147.44.84 (talk) 02:56, 21 September 2009 (UTC)[reply]

Here is how I did it:

v=at 14m/s=a(1s) 14m/s/1s=a a=14m

14m/14m/s=1s

so shouldn't the answer be 1 second?

Your "14m/s=a(1s)" line is assuming that it is travelling at 14m/s after 1 second, so of course that is the answer you are going to get. It is the average speed over the first second that is 14m/s, not the speed after 1 second. --Tango (talk) 03:04, 21 September 2009 (UTC)[reply]

but I don't know the speed of the object after 1 second, so don't I have to assume it?

The reason that you know the speed after one second is that your question said uniformly increasing. This means that the speed after one second is double the speed after half a second (and also double the average because it starts from rest). If the rate of speed increase (acceleration) is not uniform, then the calculations are more complicated. Dbfirs 07:57, 21 September 2009 (UTC)[reply]

transistor[edit]

1.What is the need of biasing a transistor? 2.What are the different types of biasing in the transistor ,Explain them in detsil? 3.What are the advantages of the biasing types ,one over the other?

Main use of biasing a transistor is for the teacher to have some questions for your homework. 95.112.150.86 (talk) 07:36, 20 September 2009 (UTC)[reply]
We have an article: Bipolar transistor biasing. See what you can learn! Graeme Bartlett (talk) 12:31, 20 September 2009 (UTC)[reply]
Most transistor equations that circuit designers want to use are actually small signal model approximations. This means that they only describe a rough approximation to the total transfer function of the device. The bias point is the region where those particular approximations are valid - usuallyalways, this means the transistor must be configured with a particular DC voltage to each input and output; the circuit equations describe fluctuations around that point. Nimur (talk) 15:28, 20 September 2009 (UTC)[reply]
Not that we should be doing homework, but Electronic_amplifier#Power_amplifier_classes is the article the OP wants. Tevildo (talk) 19:02, 21 September 2009 (UTC)[reply]

About particle accelerator[edit]

I have a question about difficulty to accelerate particles.

I have often read newspapers writing about large particle accelerators with pictures of them (though recently there seems to be not many).

Do they need a lot of energy to accelerate particles? I tried calculation, Particle accelerator#Synchrotrons says "To reach still higher energies, with relativistic mass approaching or exceeding the rest mass of the particles (for protons, billions of electron volts GeV), it is necessary to use a synchrotron." A textbook of physics at hand says 1eV = (about)1.6*10^(-19)J, so 1GeV = (about)1.6*10^(-10)J. If we accelerate 10^13 (1 trillion) protons, we need 1.6kJ.

It seems to me that facilities in pictures are too massive to provide only 2kJ or maybe 10kJ of energy. What is the difficulty in accelerating them? Like sushi (talk) 08:42, 20 September 2009 (UTC)[reply]

  • The large diameter is more to do with the Gyroradius of a high energy charged particle than some difficulty in accelerating them. Noodle snacks (talk) 09:13, 20 September 2009 (UTC)[reply]
Thank you.
Looking at Gyroradius, the reason seems to be so small charge. I can not try calculation for it, because I do not know how to handle magnetic field and what the units for it. (Could somebody do it for me?)
One more question is not the diameter, but heavily equipped tube (the one particles go through). Why is it so heavily equipped?
Like sushi (talk) 11:02, 20 September 2009 (UTC)[reply]
  • The idea here is to capture as much information about the particle interaction as possible, you would not want to miss any of the products, especially if the reaction was a rare one! Graeme Bartlett (talk) 12:30, 20 September 2009 (UTC)[reply]
On newspapers, there is usually only one picture. Is the tube covered with the heavy equippment everywhere? Or those are just one of heavily equipped places?
Like sushi (talk) 14:40, 20 September 2009 (UTC)[reply]
The synchrotron beam line has "equipment" all over it - for comparison, take a look at SLAC's beam line (which is a linear accelerator, not a synchrotron), going off about a mile each direction from this photo. Those red things are klystrons, which add energy to the particle beam; and there's a lot of coolant and so forth. Then, there are "taps" where particles are plumbed off the main beam line into special-purpose experiments. In those special regions, there is more, different "equipment" - lots of detectors, sensors, coolant plumbing, etc. Nimur (talk) 15:18, 20 September 2009 (UTC)[reply]
I was recently talking to a guy from SLAC (who had also worked at the LHC, and knows something about enormous synchrotrons). My question was about how much energy is actually stored in the "ring" at the LHC. The individual particles may reach TeV, but I was interested in the total energy in the facility. When working with large RADARs, we often use an onsite power station, because we might be consuming 3 or 5 megawatts static power, plus peaking; so I wanted to see how a particle accelerator compares. It seems like the actual energy in the highly accelerated particles is very small (e.g. kilojoules, as pointed out above); the particles themselves have long lifetimes without colliding with walls, and can basically hold that kilojoule of stored energy for hours or days. Then, for a particular experiment, the ring "bleeds off" its energy tank to shoot particles in to one of the connected experiment housings. But while in "storage" (orbiting in the synchrotron), the high-energy particles are not really needing to be "restored" with a significant amount of power. However, that's because they're in essentially a collisionless medium, thanks to the enormous electromagnets which make up the accelerator ring. These magnets, (which were the cause of the infamous fire when the magnets quenched), do require a lot of energy (I never got exact numbers and couldn't find any on the web). But, I'm told that CERN is hooked in to the local energy grid - to me, this tells me that "a lot of power" is actually less than a standard military RADAR or aviation RADAR (which often has its own electric generator). I would really be interested to see a "pie chart" breakdown of the electric consumption of SLAC or CERN - how much is used for basic facilities; how much is for the enormous computer data center(s), and how much energy is actually used for the accelerator (during a science run, of course - the accelerators typically do not operate at full power 24/7/365!) Nimur (talk) 15:13, 20 September 2009 (UTC)[reply]
The nominal annual electricity consumption for the CERN facility is roughly 1000 GWh; only 8% of that is for infrastructure (heat and lights). All the rest goes to the accelerator facilities. In addition to the actual energy required to accelerate particles, there is massive consumption of electricity required to maintain the vacuum systems (the LHC alone contains more than fifty kilometers of ultrahigh-vacuum pipe) and the cryogenic plants (the LHC holds roughly a hundred tons of liquid helium). The nominal annual consumption of the LHC alone is about 390 GWh, which averages out to about 40 MW continuous &dmash; but the beamline isn't actually on for the whole time [1]. The LHC will usually be shut down in the winter months due to high electricity costs (the Swiss use a lot of electric heat), and even during the rest of the year it will not operate at full power continuously. Based on that, I would expect a fully-powered-up LCH to be pulling 100 MW or so out of the local grid. TenOfAllTrades(talk) 17:49, 20 September 2009 (UTC)[reply]
That would be a lot of power - enough to warrant its own electric generator. 100 MW of average continuous power is about 10% of a large nuclear station's peak output capacity - enough energy for a large town or city. I suspect that if the draw is that large (and it may well be), then there must be significant on-site generation capacity. Also, thanks to TenOfAllTrade's great link, I found my pie-chart. According to that page, the electricity is jointly supplied by three separate power distributors in France and Switzerland. I imagine the technical challenge of load-balancing that quantity of power (not to mention the politics of billing it) is huge. Nimur (talk) 20:03, 20 September 2009 (UTC)[reply]
There's some more figures here [2] - it makes the point that most of the energy consumption is refrigeration. The LHC is quoted at 120MW.83.100.251.196 (talk) 20:51, 20 September 2009 (UTC)[reply]
I suppose it's impossible to properly insulate the electromagnets since that would get in the way of their interaction with the 'pipe' - can anyone say anymore about this - is it very difficult to insulate the electromagnets on the inside part?83.100.251.196 (talk) 21:01, 20 September 2009 (UTC)[reply]

what can I grow in a vase ~5" by ~30" ?[edit]

So someone unloaded a bunch of vases on me from a wedding reception. I've given a few away but am stuck with 4. They're very tall, very skinny cylinders. About 5" x 30". Ideally, I'd like to grow vegetables in them, but failing that I'd settle for a regular ol' houseplant. I've looked around a bit and there are certain species that would benefit from the depth (tomatoes...) but would really have a hard time with the width (also tomatoes...). Does anyone have any ideas? inb4 "put cut flowers in them" 74.115.162.11 (talk) 10:29, 20 September 2009 (UTC)[reply]

Drainage is going to be a problem unless you're able to make holes in the bottoms and stand them in saucers. Given that, perhaps you could grow something that requires hilling, such as leeks, gradually building up the soil as they grow.--Shantavira|feed me 10:47, 20 September 2009 (UTC)[reply]
carrots ? 83.100.251.196 (talk) 10:50, 20 September 2009 (UTC)[reply]
OP here. Thanks! They're ceramic so I can put holes in the bottom without much difficulty. I'll read up on leeks. Hadn't thought about carrots - maybe I could tie em all together and make a carrot patch! That's a lot of soil for only 4 carrots though... there's gotta be something else. 74.115.162.11 (talk) 11:51, 20 September 2009 (UTC)[reply]
The dimensions you describe sound more like they are designed to display cut flowers than grow them. I'd fill them with fully grown tall flowers or plants. Some don't even need water, like cat tails. Other could benefit from water, like lillies. However, even with water, the cut flowers will only last a few days, so just toss them, and the water, when they go limp. StuRat (talk) 12:06, 20 September 2009 (UTC)[reply]

What is WELS?[edit]

During the translation of Saint John River (New Brunswick) I have encountered this sentence: "There is an Abenaki burial site containing a large number of graves where the Big Black River joins the Saint John in township 18, range 13, WELS." The bold text should be the location of the site, however I do not understand any of it. So, please describe "township", "range" and "WELS" to me. Thank you! - Xbspiro (talk) 11:36, 20 September 2009 (UTC)[reply]

The article on Public Land Survey System has the answer. --Cookatoo.ergo.ZooM (talk) 12:09, 20 September 2009 (UTC)[reply]
Since New Brunswick is a Canadian province, you might find the answers at Dominion Land Survey. However, I would guess that the WELS term is the same as in the US, where it means "West of the Easterly Line of the State" (except that in Canada it would be the province or territory). See Wels_(disambiguation). StuRat (talk) 12:19, 20 September 2009 (UTC)[reply]
Aw, think a bit. Why would "state" be used to refer to a province? (Or territory, but there are no territories anywhere near there and never have been.) And anyway, if you read the DLS article, you'll see that that survey system wasn't used in the older parts of Canada (just as the PLSS wasn't used in the olest parts of the US), and that where it was used, things were referred to the various principal meridians, not to political boundaries.
In fact I expect Cookatoo had it right -- because the confluence of those rivres is not in New Brunswick, but in northwestern Maine, about 15 miles east of St-Pamphile, Quebec. But it wasn't the PLSS. See the Historical Atlas of the United States published in 1988 by National Geographic (ISBN 0-87044-747-5, not to be confused with any other atlas by the same name). On page 98 is a map in 6 colors showing where the different land survey systems were used, which I happen to have a photocopy of here; and for northern Maine, what it shows is "Rectangular survey, local or state".
Considering the reference to ranges of townships, I imagine that the survey used some principles similar to the PLSS, but it was not part of the PLSS. If the "easterly line" means the north-south part of the international border, then the townships cannot have been the 6-mile squares of the PLSS, as the location is less than 80 miles from that line; but maybe it refers to a line drawn north from a point farther east in southern Maine. I can't say.
--Anonymous, 22:55 UTC, September 20, 2009.
It's not inconceivable that parts of Canada near the US might share some of the same systems with the US. For example, Canada has dollars, quarters, dimes, nickels, and pennies, just like in the US. They don't use pounds or euros as in the UK. However, the sight actually being in Maine makes even more sense. StuRat (talk) 19:25, 21 September 2009 (UTC)[reply]
The PLSS plan was indeed copied in Canada as the DLS, but it was a different survey. In both countries, the act of surveying land not previously owned by anyone who counted into lots to be made available to settlers was an important exercise of sovereignty, first by the colonial powers and later by the independent countries that succeeded them. The methods of one country's survey might be copied in the other, but the survey itself would not cross the border, which was my point. --Anonymous, 19:55 UTC, September 21, 2009.
Also note that the US/Canadian border has changed at times (going all the way back to when both were colonies). Thus, if a survey was already done by one country, and that land is then acquired by another country, they might tend to go with the previous surveys rather than re-invent the wheel (and upset many locals who find their house is no longer entirely on their land). StuRat (talk) 20:20, 21 September 2009 (UTC)[reply]
Fair point. Normally one would want the border to be agreed on before surveys happen, because people also don't like to learn that their house is in the wrong country. But it doesn't always happen that way. Maine's border with what is now Canada, by the way, was defined unclearly in the original Treaty of Paris (1783) and didn't finally get settled until the Webster-Ashburton Treaty of 1842. I don't know when the areas around there were surveyed. --Anonymous, 07:32 UTC, September 21, 2009.
Thanks to your suggestions, I have found the answer in this article. Thanks again for all three of you! - Xbspiro (talk) 19:08, 21 September 2009 (UTC)[reply]
The "easterly line of the state" in Aroostook County, eh? So then they're not 6-mile square townships. Thanks. --Anonymous, 19:58 UTC, September 21, 2009.

Genetically engineered meat growing plant?[edit]

Would it be possible to genetically engineer a plant or tree to produce animal tissue for human consumption? For example a tree that grew chicken drumsticks on its branches?Trevor Loughlin (talk) 13:23, 20 September 2009 (UTC)[reply]

Theoretically it would probably be easier to genetically modify an animal that was nothing but tissue growth (you would be making an "artificial animal" of sorts), than it would be to engineer a vegetable to make animal tissue (totally different cell types, metabolisms, etc.). So for example you'd modify the DNA (and other things) of an embryo of a pig, for example, to create something that would just generate pig flesh (no head, no brain, no bones) if you put it in a solution of oxygen and nutrients, or something along those lines. It would still technically be an "animal" all the way, though it wouldn't have any nervous system (and hence no ability to suffer, and relatively simplified needs). But from a plant? I don't know—probably not. Plant and animal DNA is just way too different and probably is not in any way compatible. The closest you could probably get is plant flesh that somewhat simulated animal flesh (sort of a tofu plant) but getting something that really simulated its composition and flavor... I dunno, that's pretty hard to imagine even with very highly advanced capabilities. Much easier to just make meat from meat. --98.217.14.211 (talk) 14:11, 20 September 2009 (UTC)[reply]
A fictional treatment of 98's "artificial animal" features in Frederik Pohl & C. M. Kornbluth's classic 1953 science-fiction novel (previously a 1952 magazine serial entitled Gravy Planet), The Space Merchants. Cheap Gallina chicken meat, marketed as "Chicken Little", is harvested from an eponymous "grey-brown, rubbery hemisphere some fifteen yards in diameter . . ." - essentially an immortal cancer tumour. Yum! 87.81.230.195 (talk) 17:37, 20 September 2009 (UTC)[reply]
It should be possible to grow a food product which tastes like meat and has the texture of meat from plants, even is it is not genetically meat. See [3] for some existing plant-derived meat substuitutes. Genetic engineering might start with tofo growing from soy and seitan (sp?)growing from wheat, both very meat like. Mushrooms already yield the meat-like Quorn™. "GoodBite," made from peas and wheat, is already meatlike. Scientists already can insert animal genes into plants, per [4]. Edison (talk) 18:46, 20 September 2009 (UTC)[reply]
Nothing is impossible, but getting plants to produce animal cells is about as difficult a task as could be imagined. Basically you would have to combine the plant and animal genomes in one cell, add a "master expression controller" that switches a cell between plant and animal phenotype, add machinery to control the controller, and add some machinery that would pump sugars and other nutrients from the plant part to the animal part. At best it would take a heck of a lot of plant tissue to support even a little bit of animal tissue. Looie496 (talk) 22:45, 20 September 2009 (UTC)[reply]

Dark meat is different than light because it is excercised more. All forms of meat in a vat/grown flesh that doesn't move even a little would be the ultimate light meat and taste bad. Sagittarian Milky Way (talk) 02:19, 21 September 2009 (UTC)[reply]

Could you simulate exercise by attaching electrodes and giving the muscle tissue small shocks? --Tango (talk) 02:34, 21 September 2009 (UTC)[reply]
It worked for Galvani.. But on a plant? We got all these wires, stems, blood vessels.. it's the Borgified Frankenstein plant. Sagittarian Milky Way (talk) 02:54, 21 September 2009 (UTC)[reply]

The advantage of such a plant-animal hybrid would be that it would use photosynthesis and basic soil nutrients/fertilizer to cut out all the energy inefficiencies in the plant to animal chain, along with the labour of animal husbandry. As for exercise, the chicken legs could have a natural pacemaker to make them pulsate and also circulate blood.Perhaps a placenta like interface could convert plant proteins into blood for the muscle tissue.Trevor Loughlin (talk) 06:10, 21 September 2009 (UTC)[reply]

You're hinting at the key issue of why someone would want to do this, which is cost-effectiveness. It might be technologically possible, but if the end product isn't cheaper than raising real chickens, it would be hard to find a market for it. →Baseball Bugs What's up, Doc? carrots 06:20, 21 September 2009 (UTC)[reply]
There's no way to compete with the cost of raising chickens in, say, the current US ag environment (where the overall costs of raising the chickens in a factory setting—e.g., to the environment, or in the case of health care—are not picked up by the producer, but by the government or the consumers). One could, I guess, imagine a world where factory farming of the sort done now is illegal, and that these other costs are paid for by producers, which might make the price a bit more competitive. But that would require the actual R&D of the chicken-plant to be quite low, as the production costs are probably comparable to regular chickens... yeah, I don't know if I see it happening any time soon, to be sure. --98.217.14.211 (talk) 12:50, 21 September 2009 (UTC)[reply]

your premises are wrong, your argument is invalid, but your conclusion is correct.[edit]

any historical examples of such?

Not sure of historical examples, but argument from fallacy is related to this and contains some hypothetical examples. --Mark PEA (talk) 14:21, 20 September 2009 (UTC)[reply]
Perhaps Fermat's Last Theorem? He did say that "I have discovered a truly marvellous proof of this, which this margin is too narrow to contain" - which, given the actual proof, is probably wrong. Tim Song (talk) 14:34, 20 September 2009 (UTC)[reply]
Pretty much any failed scientific theory is an example of that.
  • Alchemy - for example - starts with the premise that the universe is made from Earth, Air, Fire and Water - proceeds to argue that (say) burning wood in air drives out the water and air and adds fire (which has negative mass in this scheme of things) - concluding that you'd wind up with some dry, Earth-like material with much less mass than the original wood (which is a fair description of wood-ash) and that you could condense water out of the gasses ("air") that are produced in the wood-smoke. The conclusion fits observations - but the premise and argument are quite utterly bogus.
  • Aristotle's theories of motion - says that when you push something, it gradually slows down and stops because being stationary is it's natural state, to which it seeks to return - and that heavy objects fall faster than light ones. This is completely incorrect - but turns out to work reasonably well down here on earth where we have air resistance and friction pretty much everywhere you look. (Dirty little secret: Several computer games use Aristotle's laws of mechanics rather than Newton's because they are easier to work with and produce pretty convincing results most of the time!)
SteveBaker (talk) 14:40, 20 September 2009 (UTC)[reply]
How about Phlogiston theory - apparently there isn't any phlogoston - but things still burn.83.100.251.196 (talk) 17:38, 20 September 2009 (UTC)[reply]
The Bohr model of hydrogen is a good one. It got the correct spectral lines despite having nearly all the details wrong. In Bohr's model the nth energy level was a circular orbit with n units of the fundamental angular momentum, but in reality the energy doesn't depend on the angular momentum at all (to a first approximation). One thing these all have in common, though, is that their predictions are really retrodictions—they were concocted to match existing experimental results. It's harder to think of totally wrong theories that successfully predicted new phenomena. I guess the Dirac sea model counts (it predicted the positron), and maybe Einstein's static universe (predicted a positive cosmological constant). -- BenRG (talk) 19:01, 20 September 2009 (UTC)[reply]
For more background on the Bohr model, the important work was done by Johannes Rydberg (see Rydberg formula) who discovered that the spectral lines of Hydrogen (see Balmer series) could be described via a simple inverse square law. That is, the wavelength difference between the lines matched, exactly, a very simple equation with simple variables. Rydberg had no idea about the electronic structure of the atom (he devised his equation in 1888, the electron would not be discovered until 1897), indeed he had no idea how his simple equation related to the actual structure of the atom. It was Bohr who connected the dots between Rydberg's equation and the organization of electrons in an atom. It turns out that Rydberg's equation is really a special case of the Schrödinger equation for a one-electron system; likewise the Bohr model only holds for a one-electron system. This is because of the problem with an n-body problem. A one electron system is a two-body problem (electron and nucleus), and as such, the interactions can be descibed by simple inverse square laws, like the Rydberg equation. Once you have more than two interactions, such as between many electrons and the each other and the nucleus, it becomes impossible to use simple mechanics to describe the state of the electrons. The relationship between the Bohr/Rydberg model and the Schrödinger model is EXACTLY analogous to the relationship between Newtonian mechanics and Quantum mechanics. As such, though it is not directly applicable in most cases, it is still a very useful model in the same way that Newtonian physics is a useful, if incorrect, model. --Jayron32 02:00, 21 September 2009 (UTC)[reply]
The Titius–Bode law is a good one - it provides a simple equation that describes the distances of the planets from the sun. It's an utterly beautiful result - and it fitted the positions of the planets that were known at the time (circa 1715) - almost perfectly. Even after Uranus was discovered - 70 years later - it fitted the rule perfectly. Sadly, Neptune and Pluto are nowhere near their predicted places - and in truth it's more or less a coincidence that the rule works so well for the inner planets. SteveBaker (talk) 21:38, 20 September 2009 (UTC)[reply]
I think it's a bad example because it doesn't meet the OP's first two requirements: "your premises are wrong" and "your argument is invalid". Being an empirical law, it had no premises; being an empirical law, the only argument supporting it was that it matched known solar system distances. The scientists hundreds of years ago didn't know why it worked, but that doesn't mean they made invalid assumptions or logical fallacies. --99.237.234.104 (talk) 23:20, 21 September 2009 (UTC)[reply]
On a similar note Kepler's Mysterium Cosmographicum gives pretty good results for planetary distances using methodology that would be indistinguisable from witchcraft to many modern scientists.83.100.251.196 (talk) 21:55, 20 September 2009 (UTC)[reply]
Sure. Even the wrong people can be right sometimes. Imagine Reason (talk) 01:13, 22 September 2009 (UTC)[reply]
Caloric theory: The premises were wrong, but the conclusions were good enough to form the basis of modern thermodynamics. --Carnildo (talk) 01:32, 23 September 2009 (UTC)[reply]
Those who espouse hidden variable theory consider the premises of quantum mechanics wrong, despite the ability of quantum mechanics to accurately predict many phenomena. However, they have yet to provide a better explanation which supersedes the quantum mechanical treatment. Nimur (talk) 21:40, 25 September 2009 (UTC)[reply]

Inexpensive LCD display that is relatively good for photo editing work — recommendations?[edit]

Does anyone know of an inexpensive (no more than $500 US) LCD display that is relatively good for photo editing work? I'm looking for one that's able to show the difference between different shades of colors near both the low end and the high end of the dynamic range.

I never realized how good my Komodo Sceptre was until I got a cheap VGA LCD monitor to put side-by-side with it. The color fidelity, contrast ratio, and overall high quality of the Komodo is fantastic, and I bought it five years ago for under $250. Your price-point will largely depend on the viewable area you want. Definitely stick to DVI (digital connection) if you are aiming for color fidelity. Nimur (talk) 15:22, 20 September 2009 (UTC)[reply]

radioactive decay: kinetics v. thermodynamics[edit]

I see with some interest (let's use isotopes of thorium as an example) that half-life has a funny correlation with decay energy. It doesn't seem to fit the Arrhenius equation model, or anything that would go with exp(). Th-231 has an incredibly short half-life, even though the decay energy is only 0.39 MeV. The decay energy for Th-232 is 4.0 MeV, though it has a half-life of 11.4 billion years, but the half-life of the next most stable isotope is six orders of magnitude smaller, even though the decay energy only decreased by 0.7 MeV. John Riemann Soong (talk) 16:28, 20 September 2009 (UTC)[reply]

Nuclear decay requires a quantum mechanical treatment, properly adjusted for the large number of interacting nucleons. Have you analyzed the decay from that point of view? Classical atomic theory and even the simplifications of quantum theory that work well for "electron clouds" are rarely rigorous enough to solve the nuclear reactions such as fission and decay. Nimur (talk) 16:57, 20 September 2009 (UTC)[reply]
The decay ernergy wouldn't be the right quantity anyway - the Arrhenius equation uses activation energy not reaction enthalpy.
You'd need the energy of the transition state between starting nucleotide, and deacy products, and as Nimur says a classical treatment might not still be good - at the very least I'd would expect something like the equivalent of quantum tunnelling to be important (since I expect the nucleus to be very small)
That said there might be a classical relationship between energy of decay, and the activation energy (and hence half life) - however this would be complicated by the structure of the nucleus... (what ever that is exactly)
If you want to try a classical comparison that might work - I'd suggest comparing half lifes of nucleotides differing by 2n+2p (since I already know that that is a significant number..)
It's a bit like comparing properties of transistion elements - just as you need to take into account electron configuration for those, you'll probably need to take into account (or look for patterns) in nuclear structure as the atomic number increases..83.100.251.196 (talk) 17:35, 20 September 2009 (UTC)[reply]
Leaving aside the quantum mechanical considerations for a moment, the trick to using the Arrhenius equation is that Ea is defined as the activation energy of the reaction in question. The decay energy that you're looking at is related to the net energy change of the reaction; it doesn't tell you anything about the activation energy of the decay 'reaction'. What the change in energy will tell you is the ratio of reactants to products (as a function of temperature) after the system reaches equilibrium; see Boltzmann distribution. TenOfAllTrades(talk) 18:02, 20 September 2009 (UTC)[reply]
Almost forgot - are you comparing the same decay processes - there are a lot of different types beta decay, alpha decay etc that would complicate an already complicated relationship.83.100.251.196 (talk) 20:33, 20 September 2009 (UTC)[reply]

Victoria Falls swimmers[edit]

Tourists swimming at the edge of Victoria Falls

htf is this possible? — Preceding unsigned comment added by 86.162.106.217 (talk)

The brownish area between the swimmers and the falls is a rock lip just beneath the surface. See Snopes for more info. DMacks (talk) 19:22, 20 September 2009 (UTC)[reply]
The image description page states "It is possible to swim at the edge of the falls in a naturally formed safe pool, accessed via Livingstone Island." and from our article on Victoria Falls "A famous feature is a naturally formed pool known as the Devil's Armchair, near the edge of the falls, accessed via Livingstone Island. When the river flow is at a safe level, usually during the months of September and December, people can swim as close as possible to the edge of the falls within the pool without continuing over the edge and falling into the gorge; this is possible due to a natural rock wall just below the water and at the very edge of the falls that stops their progress despite the current." Nanonic (talk) 19:25, 20 September 2009 (UTC)[reply]
That does not mean that these tourists are participating in a "safe" activity. Nimur (talk) 19:58, 20 September 2009 (UTC)[reply]
It's dangerous, just not as dangerous as it looks. Although, given the way waterfalls work, that rock lip could shatter at any time. For a more daring similar illustration, go to this NG page [5] and scroll to the right about three times. Look for a thumbnail of an African in silhouette. That's dangerous. Baseball Bugs What's up, Doc? carrots 20:24, 20 September 2009 (UTC)[reply]
Took me ages to realize you weren't talking about the baby playing with cobras. ~ Amory (usertalkcontribs) 22:32, 20 September 2009 (UTC)[reply]
Well, that doesn't seem too safe either. But I was referring to the photo of the African standing on the lip of that pool. One little slip and you'd be fish food... after a nice 30-story drop. Baseball Bugs What's up, Doc? carrots 22:36, 20 September 2009 (UTC)[reply]
(According to the text, the cobras are defanged. So that's not sooo bad...) --98.217.14.211 (talk) 12:55, 21 September 2009 (UTC)[reply]

Declining bee population?[edit]

Human carnivorous behaviour has reduced flora/plants and in turn is threatening the bee population. Is this true? I don't read it on Wikipedia and I don't know if other sources are good or just there is two of everything on the internet... Anybody know about bees? ~ R.T.G 20:21, 20 September 2009 (UTC)[reply]

Sort of this was in the news in the UK [6] [7] [8] - there seemed to be an unexplained number of colony collapses Colony collapse disorder which may be unrelated to farming methods.
Excluding the potential for a bee virus, increased levels of monoculture, lack of wild flowers, various insecticides, etc are a problem for bees - topics briefly mentioned in the above newspaper links.83.100.251.196 (talk) 20:27, 20 September 2009 (UTC)[reply]
I think we should be linking this colony collapse disorder to bee because in Ireland there is an apparent 50% decline and according to some here that only follows a similar happening in America and now from this we see it is a world phenomenon. Thank you. ~ R.T.G 20:32, 20 September 2009 (UTC)[reply]
CCD is linked to bees, and vice versa. What link do you think is missing? --Tagishsimon (talk) 18:03, 21 September 2009 (UTC)[reply]
Colony collapse disorder does mean "declining bee population." It is a specific type of loss of bees. Most claims of colony collapse disorder turn out to be something different. If you read the article colony collapse disorder, it details that much of the "world-wide" claims of CCD are not in any way CCD or still unsubstantiated. -- kainaw 20:38, 20 September 2009 (UTC)[reply]
I don't know what you mean but the article seems to confirm that bees are mysteriously dying in large numbers around the world? ~ R.T.G 20:40, 20 September 2009 (UTC)[reply]
The CCD article makes is painfully clear that many cases that are initially classified by farmers as CCD turn out to be population loss for another reason. Therefore, your request that the CCD article be linked directly to the main bee article because it is a "world phenomenon" requires some substantial evidence that it is a world phenomenon. I think you are using CCD as a synonym for "population loss". CCD is one cause of population loss. There are many others that are not CCD. -- kainaw 21:27, 20 September 2009 (UTC)[reply]
I have to disagree. Much like the term AIDS was used to mean any unexplained deficiency in the human immune system, prior to the identification of an agent (HIV, in this case), CCD is currently used to mean any unexplained loss in bee population. Many possible agents have been identified, in this case, so CCD really doesn't yet have a specific meaning. StuRat (talk) 17:16, 21 September 2009 (UTC)[reply]
And beside that, going by the article claims all we would need is incidents in China and we would have circled the globe. Will that do for a broader implication of "worldwide"? Seriously it says Canada, 25 USA states, 7 or 8 European countries and suspicious reports from Brazil and India, that is quite considerable and completely overlooked on the bee article as well as by the experts here on the "everything dies if the bees die" ~ R.T.G 17:51, 21 September 2009 (UTC)[reply]
That's overstating it a bit. Everything wouldn't die even if bees all went extinct (which is extremely unlikely, as they will mutate and adapt to handle whatever threatens them). There are other pollinators for flowers, such as birds and other insects. As far as people are concerned, we can manually pollinate plants (or possibly build machines for that purpose), or we can eat foods that don't require pollination (such as those which reproduce asexually). Life goes on. StuRat (talk) 19:13, 21 September 2009 (UTC)[reply]
Sorry Stu but that is the naive view. Is it fair for us to say "They'll mutate and cope with it!"? Well, I think the problem must be a bit more serious than that! ~ R.T.G 01:19, 22 September 2009 (UTC)[reply]
No, it's not naive at all. Insects have very short lifespans and high mutation rates, allowing them to develop resistance far more quickly than us. So, while it took humans centuries to develop resistance to the Black Plague, bees could likely do it in a few years. StuRat (talk) 16:03, 22 September 2009 (UTC)[reply]
If anyone has any citations that state CCD has lost any and all meaning and currently means "any reason for a bee to die", then please update the article. Otherwise, I will continue to stick with the cited article over claims otherwise. CCD has a very specific meaning. As the article states: "a major part of the subsequent analysis of the phenomenon hinges upon distinguishing between true CCD losses and non-CCD losses." What is happening in this discussion is a rationalization that any loss is CCD. We might as well claim that cars cause CCD when they hit bees while driving down the road. I sincerely hope that the absurdity of such a claim is obvious and, I hope, makes it clear that making up your own definition of CCD is just as absurd. If you want to show concern over declining bee population, that is fine. Trying to redefine CCD to be anything you want it to mean takes away from the message and makes it sound absurdly alarmist. -- kainaw 04:01, 22 September 2009 (UTC)[reply]
You changing the def of CCD from "any unexplained death of a bee colony" to "any death of a bee" is a straw man argument. Nobody is arguing that an explained cause of death, like a car hitting a bee, is CCD. When you refute an argument, please stick with the actual argument, and don't just exaggerate it to make fun of it, that's not at all helpful. StuRat (talk) 16:03, 22 September 2009 (UTC)[reply]
Please, do not detract from the important factor one more time. CCD? Who cares. The bees are dying. There is speculation that such a thing would be frighteningly dangerous to life as we know it. There is not so much as a scare mongerer on the topic to inform us about that problem. The closest we have on Wikipedia, apparently so far, is the CCD article. Defining CCD is a secondary almost irreleveant topic unless you have zero real interest in the important topic, the bees are mysteriously dying all over the place and nobody seems to know about it. It is so unheard of, I imagine people saying "You made that up". If you can think of a more relevant subject than CCD please share with us. If you want to discuss disparity between language and expert speculation on CCD, try Talk:Colony collapse disorder. I can see no topic more worthy of linking to the article Bee wether they aren't sure if it was CCD or not unless it is a hoax article? ~ R.T.G 12:52, 22 September 2009 (UTC)[reply]
It appears that you now understand my point: CCD is a type of mass-death in a bee hive. It is not the term used to define any and all mass-deaths in a bee hive. Understanding that point, I hope you understand why I disagree with your claim that we should "link" CCD to bee. It may very well be that your definition of "link" is different than mine. It appears to me that you are claiming that we should take the article on bees and replace it with the article of CCD. My only intention in this entire thread has been to point out that CCD is a type of mass-death for bees. It is not a definition of bee. Therefore, if someone wants to learn about bees, getting an article about CCD will not be very helpful. If you mean something completely different by "link", please explain yourself further. -- kainaw 19:25, 22 September 2009 (UTC)[reply]
Sure Kainaw, the article about bees dying is completely irrelevat to the subject of bees. I am glad, in a way, that you can see how I understand you because my OP would not have been answered satisfactorily if only for you perhaps, wouldn't you say so? Like you say, if learning about bees, knowledge of CCD would not be very helpful even though it was the most useful addition to this thread so far, we will take your word for it until you add something better, right? Oh, I can see you understand me too so that is all settled, yay. ~ R.T.G 10:00, 23 September 2009 (UTC)[reply]

In true American fashion, Yanks assume that because they have just about destroyed THEIR OWN bee colonies through rapacious greed, the same is true EVERYWHERE else. It is not. In Australia, our bees are disease-free, healthy, happy, and producing the world's best honey. But then we don't drive them around in trucks for thousands of miles, forcing them to fertilise different fields day after day, and thus multiplying over and over again their chances of contracting a disease, and making their lives unbearable because they have to learn new geographies and new flowers just as they become familiar with the old ones. Nor do we force them to eat stuff from genetically modified foods. And in reward, by letting them feed on the pollen of a particular flower at their own leisure, our contented bees produce honey with the exquisite taste and aroma of THAT flower, and other countries beat a path to our door to buy jars of it. The destruction of American bee colonies, with genetic diseases leading to malformed bees, and the collapse of their immune systems through being forced to exist in highly unnatural slave type conditions has been well documented. Myles325a (talk) 05:06, 26 September 2009 (UTC)[reply]

Hmm Australia? You mean a possible source of Diseases of the honey bee#Israel acute paralysis virus (IAPV) which has been linked to CCD? [9] (yes I'm aware this raises doubt on the claim Australian bees is what brought it to the US) [10] [11] [12] Incidentally, I doubt anyone forces bees to 'eat stuff from genetically modified foods' they choose to do it themselves. Also the idea this is solely an American problem doesn't explain why it's a problem in parts of Europe (as mentioned in the CCD article and in the Wired article), which don't have many of the causes you mentioned above. Incidentally I'm doubtful of the claim Australian honey is the world's best. Jellybush honey aka Australian Mānuka honey is not as well known as (real New Zealand) Mānuka honey, in fact it isn't even mentioned in Monofloral honey and the fact that it's sometimes called Australian Mānuka honey should given an idea too. Nil Einne (talk) 07:55, 26 September 2009 (UTC)[reply]
This sounds like the same anti-American rhetoric I'd heard when AIDS was first documented in the US (note that it had apparently existed for years in Africa before that, without being documented). People in other countries argued that they didn't have to do anything about it since "it's just an American disease". Now it's a serious threat in many of those nations which chose to ignore it. StuRat (talk) 12:01, 26 September 2009 (UTC)[reply]

Myles325a back here. Sturat, argument by analogy is a poor way to discuss anything, and is not counted as evidence. Nil Einne, yes there have been attempts to blame Australian bees for American bee problems, but informed sources have discounted that, and many put forward the idea that such a slander has been promoted to avert attention from the rapacious practices that have led to the decline of the bee in the US. How is it that even now there appears to be NO clear answer as to what has caused these major epidemics, so dangerous to the fertilisation of crops? Could it be that big business has hushed it all up to protect their own pockets? Your other comments are amongst the silliest I have read here for quite a while (and that's saying something.) "Bees choose to [eat what they eat] themselves", do they? Yeah, my cat doesn't eat meat...I don't give it any. Bees are not window-shopping credit card carrying flits from Sex and the City. When they are put into some huge field with identically cloned and genetically modified sunflowers, they have to eat that, don't they? And Australia has hundreds of differnt honeys. How does asserting that ONE brand of honey (Manuka) is actually a New Zealand honey go towards rebutting my claim that our honey is the best in the world? Of course New Zealand's is 2nd best, and I wouldn't eat the stuff from your place if they tied me down and poured honey down a funnel wedged in my throat. Myles325a (talk) 23:06, 27 September 2009 (UTC)[reply]

Long time later, but my point is that Mānuka honey is arguablly far more likely to be considered the best honey then anything from Australia. There's good evidence from this from the fact monofloral honey mentions 3 monofloral honeys from Australia. One of them is Mānuka honey which as I said is generally considered more of a New Zealand thing and is well known enough to have an article. Another is Eucalyptus where the article barely mentions the honey. A third is a subtype of the former Eucalyptus marginata where again the article barely mentions the honey. Incidentally, it sounds like you don't know how science works, but from your history, I guess this isn't surprising. Your reply BTW is pretty silly. My cat eats meat sometimes it will even eat meat I leave lying around if I'm careless. What it will not eat (unlike dogs) is chocolate probably because it can't taste sucrose. The only way I could get it to eat chocolate would be either by really forcing it. Giving it meat to eat isn't forcing it to eat meat, just the same as putting bees in a field containing GM crops isn't forcing them. It's simply what we give them. The use of the word 'force' implies an unnecessary degree of coercion which isn't suited to the realities of the situation. Bees will eat whatever is similar enough to what they've evolved to eat that's in the environment. Incidentally, you still seem to be ignoring the fact that CCD is not a problem unique to the US. Nil Einne (talk) 08:48, 5 July 2012 (UTC)[reply]

Jupiter's gravity protecting Earth from comet impacts[edit]

I've read that Jupiter's gravity protects the Earth from comet impacts. How does this work? --Halcatalyst (talk) 21:20, 20 September 2009 (UTC)[reply]

Jupiter is big and has a large gravity field. As a result, it's got a good chance of interacting with passing comets such that they either get flung out of the solar system or, more rarely, collide with Jupiter directly. It's also possible that Jupiter's gravity tweaks a comet just so and causes a collision with Earth, but those odds are substantially lower given that the Late Heavy Bombardment is past and most remaining space bits are in stable orbits). To a lesser extent, the other gas giants provide similar protection, but they're both less massive (thus less gravitational influence) and farther away (thus covering proportionally less of the solar system). — Lomn 21:43, 20 September 2009 (UTC)[reply]
The idea being that over time Jupiter will vacuum up or toss out many of the random space objects in orbits around the sun that cross Jupiter's orbit. Those same objects tend to be the ones that could potentially hit Earth. Sometimes the interaction between Jupiter and a comet might just redirect the comet to some other orbit, which on average probably doesn't hurt or help our chances of being hit, but when a comet collides with Jupiter or gets flung out of the sun's orbit, then we have one less comet in the solar system to worry about. Rckrone (talk) 22:07, 20 September 2009 (UTC)[reply]
It's of note that at Jupiter#Impacts it says that actually it's unclear whether Jupiter actually does this. --98.217.14.211 (talk) 00:19, 21 September 2009 (UTC)[reply]

I too think this idea is one that sounds straightforward but is less so when you give it more thought. Sure, comets and other stuff will collide with Jupiter and thus be mopped up, and no longer be a danger. And its immense gravity will make other coments deviate from their paths. But wouldn't such deviations be divided between those comets which are expelled from the Solar System and those which are propelled inwards where they are more of a danger to Earth than ever? We might expect that many comets' orbits are around the sun. If those orbits are very eccentric, they will pass the Earth again and again, each time from a different direction and by a different distance. This is what Halley's comet does now. And eventually the comet could hit Earth. A large gas giant like Jupiter could have the effect of disturbing that orbit on one of the comet's visits to our Solar System. That disturbance could well mean the connection between the comet and the Sun was changed and there would be a good chance that it would be thrown out of the sytem. That would be better for the Earth. In the early days of the Solar System, there was a lot of debris around, and we can see by the Moon's surface that there were a lot of collisions. It is thought that our moon was formed after a major collision between the Earth and another body ( which also gave the Earth its tilt). Some planets have been knocked sideways by immense collisions. But over time, these potentially dangerous bodies have been either mopped up, herded into roughly circular orbits or thrown out of the System. Planets like Jupiter may have been quite important for this process. Incidentally, it has been proposed that the Earth has had additional protection from the moon which circles it, in a local and smaller version of what Jupiter is thought to do. Myles325a (talk) 00:44, 21 September 2009 (UTC)[reply]

Two points. First: Comet Halley's orbit doesn't come anywhere near Earth's orbit. Second: Comets sent by Jupiter into the inner solar system are likely to have a second encounter with Jupiter later on when it might be ejected. If it doesn't get ejected then it is likely to have a third encounter and than a fourth one , etc... On the other hand, once a comet gets ejected it never comes back, and that's the point of the "Jupiter mops up comets" theory. Dauto (talk) 01:13, 21 September 2009 (UTC)[reply]
To draw an analogy, its like cars driving on a snowy road. As the tires scatter the snow, some of the snow lands somewhere else on the road, where the next car can still drive through it. However, any snow that gets thrown from the road will never have any mechanism to get back on the road, so it is gone forever. As hundreds of cars drive along the road, it will gradually clear of snow, since each car clears some of the snow. Likewise, Jupiter does not actually have to clear every comet the first time it encounters it. It only has to clear it from the solar system once; and Jupiter has had millions of orbits to "clear the snow" as it were. Even if there is only a tiny chance on any one encounter that Jupiter will eject any one object, after many thousands of such encounters, they all become cleared out. --Jayron32 01:42, 21 September 2009 (UTC)[reply]
You are ignoring the fact that nearly all comets crossing the inner solar system are "new", meaning that they were scattered into the inner solar system due to Oort cloud and interstellar interactions within the last few million years. So, in our mature solar system, the population of inner solar system crossing comets is a balance between processes that scatter comets into the inner solar system (from the practically limitless population beyond Neptune) and processes that destroy or scatter them out again. It is not the case that Jupiter could ever destroy all the comets. Dragons flight (talk) 02:05, 21 September 2009 (UTC)[reply]
Halley's Comet has an aphelion of 35.1 AU and a perihelion of 0.586 AU, so it must cross the Earth's orbit occasionally (every few thousand orbits, maybe - its orbit is highly inclined compared to Earth's orbit, so we have to wait for them to precess so the nodes coincide). A collision is extremely unlikely, but not impossible. --Tango (talk) 01:55, 21 September 2009 (UTC)[reply]
Given that we passed through Halley's tail in its 1910 sighting, a collision with that one is certainly within the realm of possibility. Baseball Bugs What's up, Doc? carrots 02:56, 21 September 2009 (UTC)[reply]
The tail is literally millions of times longer than the solid part and thousands of times wider. No big iffy. Your death has only a 1 in 25000 chance of being by comet or asteroid. Sagittarian Milky Way (talk) 03:10, 21 September 2009 (UTC)[reply]
No question about it. But it's still possible. Millions of miles sounds like a lot, but it's a blip in interstellar space. It came relatively close once. It might be millions of years before it comes that close (or closer) again. But it could happen. Baseball Bugs What's up, Doc? carrots 03:37, 21 September 2009 (UTC)[reply]
Tango, as you said, Halley's orbit has a high inclination and that's why its orbit does not come anywhere near earth's orbit as I had originally stated. That's all I said. Slow processes such as chaotic orbit precession could conceivably over a large amount of time change that orbit into a different one where the comet might get closer to earth's orbit, but that is impossible to predict with any certainty. Dauto (talk) 15:57, 21 September 2009 (UTC)[reply]

Masturbation[edit]

Removed request for a medical prognosis based on a change in activity. -- kainaw 22:21, 20 September 2009 (UTC)[reply]

To the person who asked the question: if you ask for facts without relating them to yourself, you may have better luck getting answers. And you should start by reading masturbation. Looie496 (talk) 22:25, 20 September 2009 (UTC)[reply]