Wikipedia:Reference desk/Archives/Science/2009 February 25

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.

February 25

Pipe noise

If I turn on the hot water in my shower with limited flow (not a trickle, but not a gush), within a minute the flow slows to a trickle and it starts making a rapidly-repeating banging noise that doesn't stop until it's turned off or the flow is increased. Cavitation, water hammer, or something else? arimareiji (talk) 00:13, 25 February 2009 (UTC)

Yes, mine too! A whistle, indeed. What is it? I guess that that particular speed of the flow corresponds to a proper frequency of the tap and make it resonate. --pma (talk) 00:32, 25 February 2009 (UTC)

Usually, but not always, it's just something stuck in your aerator (faucet) or showerhead. Unscrew that and descale it with some vinegar. See if the noise persists. 76.97.245.5 (talk) 01:03, 25 February 2009 (UTC)

That jogged my memory. What you describe happens when the screw holding the washer works loose. I've actually seen that, the prohibition against original research be damned. --Milkbreath (talk) 05:09, 25 February 2009 (UTC)

This little phenomenon occurs to me as well. The banging noise is called Water hammer, and if it starts, you should stop it quickly by turning it off/increasing the flow, whatnot. The signature bang characteristic of water hammer is caused by rapidly increasing pressure in your pipes. If it continues, your pipes could burst from the pressure, so I would remedy this with whatever remedies happen to appear here. Until It Sleeps 05:24, 25 February 2009 (UTC)

More OR, recently the plumber changed the washer and ended 18 months of water hammer. Ya-aayy! Julia Rossi (talk) 12:53, 26 February 2009 (UTC)

Silver nitrate storage

why is AgNO3 kept in coloured bottles??? —Preceding unsigned comment added by 123.50.168.74 (talk) 13:15, 25 February 2009 (UTC)

Although it is, as our article says, "relatively stable to light," keeping light out of the containers it's stored in minimizes the breakdown of the compound. Deor (talk) 14:12, 25 February 2009 (UTC)

Because it is sensitive to light. If it is exposed to too much light it turns into silver (I think). As our article says, it isn't as sensitive as silver halide, but it is still sensitive. --Tango (talk) 14:15, 25 February 2009 (UTC)

You're right about elemental silver being a decomposition byproduct. I think the photodecomposition reaction is the same as the thermal decomposition reaction which is 2AgNO₃ ---> 2Ag + O₂ + 2NO₂ 152.16.253.109 (talk) 18:49, 25 February 2009 (UTC)

Silver nitrate used to be used for photographic work - so it is indeed sensitive to light. But the main reason it was replaced is because it's not sensitive enough for dim light or short exposures. So it's RELATIVELY insensitive (compared to other photosensitive substances) - but a dark glass jar is good enough. SteveBaker (talk) 16:39, 25 February 2009 (UTC)

Actually, it's because it happily forms silver oxide if the jar isn't completely airtight and humidity gets inside, or if it gets opened a lot. If you work in a chemistry lab with a solution of it and it touches your bare skin, you'll soon know because it turns an ugly brown/black as soon as it's exposed to sunlight. Fortunately, it wears out of your skin in a week or so.

Which is a lot better than what happens to people who try to medicate themselves with silver ions - they can turn permanently blue (argyria). No, I'm not kidding. arimareiji (talk) 22:50, 25 February 2009 (UTC)

Clear glass would stop air or moisture getting in. Coloured class is used because of light. --Tango (talk) 00:01, 26 February 2009 (UTC)

high electric tension

Hello I heard that when there is very high electric tension (eg 1000000 Walt), there is no longer a risk to humans. 1 Is it true? 2 What is the explanation for this phenomenon?

In the article Electric shock it is written that the danger is primarily in the tensions between 30 and 250 volt, it may be related to the subject.

Thank you very much —Preceding unsigned comment added by 82.81.19.138 (talk) 13:26, 25 February 2009 (UTC)

I think the reason high tension cables aren't dangerous is because they are kept up high out of the way so you can't form a complete circuit to ground. High voltage means lower current, which might make a difference, but as electric shock from a high tension cable will still kill you. --Tango (talk) 14:19, 25 February 2009 (UTC)

There may be some confusion, as electric tension is a measure of voltage, not wattage (electric power). A high-tension power line will be VERY dangerous - "lower current" does not mean "safe levels". A Van de Graaff generator, on the other hand, is high voltage and very low current, and is generally not harmful (but this is not a guarantee - you can still get a painful shock, depending on what you do with it!) Finally, high frequency electricity will be subject to skin effect and will probably not penetrate deeply into the human, making it "safer" than lower frequency AC current. Again, none of these are guarantees of safety - there are situations where skin-effect is not sufficient to protect you. In the case of very high power (watts), there are other safety issues, including thermal heating, which may be equally dangerous as electrocution. Nimur (talk) 16:06, 25 February 2009 (UTC)

I think there is a fundamental confusion here. The risk due to touching - or even getting within a few feet of a high tension power cable is most definitely very high indeed. What I think the OP is referring to is that it is (perhaps) no longer considered dangerous to live, work or generally be close to high tension power lines. There has been much controversy over this in the past where people who live or work close to these gigantic pylons have complained of a variety of long-term medical conditions that have been loosely attributed to some aspect of those power lines. There has been more talk of this in the media quite recently because the 'Obama stimulus package' includes a boatload of cash for creating a much improved electrical grid throughout the USA in order to allow far distant wind and solar power plants to distribute their power all across the country - even though they are (by necessity) situated in fairly remote an/or inhospitable places. This huge build-out of new high tension power lines will doubtless upset a great many people - and the debate as to whether they are safe or harmful over the long term has resurfaced. However, there is absolutely no doubt that if you get within some short distance of the cable itself, you could come to a great deal of harm! SteveBaker (talk) 16:36, 25 February 2009 (UTC)

The prime element in the lethality of electricity is the internal resistance of the voltage source. If internal resistance is high, only a small amount of current can flow in an external circuit. It is current that harms a person. A high voltage can impose a shock, but will not cause harm if the internal resistance of the voltage source is high. There are other factors involved, such as the amount of sweat on the skin, place of contact, etc. But internal resistance is the prime element. The internal resistance of a voltage source is often overlooked. For instance, the Wikipedia articles "Cattle prod", "Electric shock", "Electroshock weapon", and "Taser" make no mention of it. Also, electronic discussions in general often ignore the internal resistance of a voltage source when it should be mentioned. – GlowWorm. —Preceding unsigned comment added by 98.16.66.64 (talk) 17:25, 25 February 2009 (UTC)

Internal resistance is very rarely high enough to avoid electrocution! Dbfirs 17:40, 25 February 2009 (UTC)

In reply to the OP:

1. No, it is not true. There are a few special types of very high-voltage sources that are not always lethal to humans (like this), but that in no way justifies the general claim that 'there is no longer a risk'.

2. See 1.

Electric shock: the statement in that article is unhelpful, but it probably means that more people get electrocuted by 110 to 250 volts because that is the voltage range that is supplied to practically every building in the developed world. That's why these voltages cause more deaths than any others.

--Heron (talk) 18:41, 25 February 2009 (UTC)

Could someone with CIPA still die from a chest stab?

(First off, why isn't there a Medical section of the RD?)

If someone with CIPA (the inability to feel pain) got stabbed in the heart, would they still live since there's no pain to kill them? --129.130.239.68 (talk) 15:50, 25 February 2009 (UTC)

Pain isn't lethal, but bleeding (from being stabbed in the heart, say) can be. I'm dubious that someone who can spell "anhidrosis" doesn't know that. --Sean 15:55, 25 February 2009 (UTC)

It was pasted, not spelled, Sean. --129.130.239.68 (talk) 16:10, 25 February 2009 (UTC)

There isn't a medical section of the RD because we aren't allowed to answer medical questions. General questions about biology (which you might reasonably describe this one as) belong here - with the other science questions. We generally resist the temptation to add new RD sections because we only have a fairly fixed number of respondants and increasing the number of desks doesn't really change that...if anything, it makes matters worse because someone who is not generally interested in answering 'medical' questions wouldn't see your question because they wouldn't patrol that desk. Putting the question here in science guarantees the most 'eyes' on the problem. SteveBaker (talk) 16:27, 25 February 2009 (UTC)

People or animals under general anesthesia, or unconscious due to lack of oxygen, also do not feel pain, but can die from blood loss associated with stabbing or cutting, shooting, decapitation, or other Wounds, as the original questioner might realize on a bit of reflection.The article Death may be helpful in understanding how Life processes in animals can be interrupted even without the experience of Pain. Edison (talk) 17:44, 25 February 2009 (UTC)

Pain is not what kills you when you are stabbed in the heart, generally. Blood loss, probably blood pressure dropping, things like that—that's what kill you. Whether you feel it or not hardly matters except in cases when you even have time to go into shock—I suspect getting your heart stopped is not one of those cases. --140.247.243.27 (talk) 20:59, 25 February 2009 (UTC)

Technically, it's actually because:

1. Your heart is enclosed in a tough protective sac called the pericardium.
2. If a wall of your heart is punctured, every time it tries to pump blood through your body it squirts some into the pericardium.
3. It doesn't take long for the pressure from the blood outside the heart, in the pericardium, to make it impossible for the heart to pump blood.

And as the worthies before me have noted, this is completely unaffected by the ability (or lack of ability) to feel pain. arimareiji (talk) 23:08, 25 February 2009 (UTC)

How can you stab a hole in the heart without also putting at least one hole in the pericardium ? So, then, why would pressure build up there, as opposed to leaking out of the pericardium into the abdominal cavity or completely out of the body ? StuRat (talk) 19:19, 26 February 2009 (UTC)

Spot on, Stu. Arimareiji was trying to describe cardiac tamponade, which most commonly occurs because of pericarditis with a fluid effusion, and from causes of cardiac bleeding, such as rupture of a coronary artery, such as might occur as a result of atherosclerosis, or from rupture of a ventricle wall, such as might occur following a myocardial infarction or as a consequence of a ventricular aneurysm. The atria of the heart don't tend to rupture because they are low-pressure chambers. Tamponade may also arise from other conditions leading to fibrosis of the pericardium. If you've been stabbed in the chest, then the things that will cause you to die are 1)blood loss due to perforation of the heart, great vessels (aorta, pulmonary trunk etc) and/or lungs; 2)Hypoxia, as a result of either 2a)transection of the trachea, or either bronchus in the mediastinum; 2b) interruption of blood supply to the lungs. Furthermore, any mediastinal or lung injury will cause either a pneumomediastinum, haemopneumomediastinum, pneumothorax, or haemopneumothorax; any of these can go on to become the "tension" variety, which causes a mass effect in the chest, which exacerbates either the tamponade and/or respiratory compromise, hastening death. No-one ever died from pain alone! Mattopaedia (talk) 07:17, 27 February 2009 (UTC)

It does seem like pain could cause death, via the following mechanisms:

1) Acute pain and the associated stress could trigger a heart attack, stroke, or other fatal events.

2) Chronic pain and the associated stress could trigger or worsen an existing medical condition, leading it to be fatal where it might have otherwise been survivable.

3) Chronic pain could cause the individual to choose suicide. StuRat (talk) 15:46, 27 February 2009 (UTC)

...Time to reset the indent...

Well, no. You see, all of those are indirect causes of death associated with pain:

1. Pain is the spinocerebral response accompanying nociceptive stimuli. In other words, pain is the aversive signal you get when unconscious signals tell your brain some part of you is being damaged. It is not damaging in itself.
2. Pain is subjective. All people percieve pain differently for a given stimulus. Our article on pain even suggests there are heritable factors related to this.
3. We all respond differently to pain.
4. Our responses, which are largely conscious decisions and therefore largely modifiable, have a significant bearing on the consequences of the pain:
   1. If one chooses to panic, then one will experience emotional stress. If it happens one also has ischaemic heart disease, a cardiomyopathy, or a vasospastic disorder like Prinzmetal's angina, the result of the physiological effects of the stress (the adrenergic effects of the fight or flight response) could induce enough transient ischaemia in the myocardium to result in myocardial infarction, or could elevate blood pressure enough to rupture a weak cerebral artery and cause a haemorrhagic stroke.
   2. If one chooses not to seek appropriate assistance to learn to develop effective coping strategies for their chronic pain, then one may develop a reactive or situational depression, which, if untreated, can lead to attempting suicide.

In each example you cite it's not the pain that's primarily responsible for death. It's associated, sure, but ultimately it's our response that determines the consequences. Hence my earlier assertion: no one ever dies from pain alone. Mattopaedia ^{Have a yarn} 02:43, 28 February 2009 (UTC)

You seem to be assuming that pain management strategies which are 100% effective are available to all. I doubt that they are. Thus, some who seek help will still suffer from the consequences of pain, possibly including death. I also forgot to mention a fourth way that pain can kill:

4) Pain can make exercise difficult, and thus can lower the level of fitness of an individual, leading to health problems like obesity, which may eventually cause death. StuRat (talk) 14:31, 28 February 2009 (UTC)

Thanks for your interest, Stu. I'm not sure how you came to the conclusion that I've assumed pain management strategies that are 100% effective are available to all. I never said that. I never implied that either. Just because there's no chronic pain service in your town doesn't mean its not available either. In Australia for example, people who live in rural and remote communities may have to travel hundreds of kilometres to access health care services that are considered basic in cities. Pain management facilities exist, but you may have to choose to travel to access one. I have some specific comments about pain management, based on years of clinical experience. Firstly: the only pain management strategy that approaches 100% efficacy is general anaesthesia. Secondly: all pain management strategies are variably partially effective (different effects for different people, different effects in the same people over time). This includes both drug and non-drug methods. Thirdly, and this especially applies to chronic pain management: the success of any pain management strategy depends on the sufferer being in a mental state that allows them to deal (ie cope, or live with, or suck it up) with the pain they have, maintain as much of a normal life as they can, and be as positive as they can. That's pretty much the core of pain management psychology. If the individual is not successfully able to maintain this mindset at all (or even some of the time), then my previous comments regarding reactive depression and suicidality may apply. Now to your fourth option. Yes, pain can make exercise difficult. But not necessarily impossible. Any kind of movement will burn kilojoules, if you choose to do it. Even walking to the fridge! Obesity is also preventable by making better choices about what & how much you put in your mouth. It's still the consequences of our actions or responses to pain that can kill, not the pain itself. Cheers, Mattopaedia ^{Have a yarn} 02:37, 2 March 2009 (UTC)

I'd bet that if you did a study of people who are otherwise in identical health, that those suffering from acute or chronic pain would have lower life expectancies. This would be because, even with actions taken to control pain and it's effects, people in pain are still more likely to die as a result of the reasons I listed, than those who suffer no pain. Thus, it's appropriate to ascribe pain as the primary causative factor in those additional deaths. StuRat (talk) 15:13, 3 March 2009 (UTC)

There's been some studies done of all-cause mortality in chronic pain. A quick Google of "chronic pain mortality" turned up these in the first page: [1] [2] [3] [4]. One article did not find a significant increase in all-cause mortality in chronic pain suffers versus normal control subjects. The other articles did find a significantly increased risk of all-cause mortality in chronic pain suffers, one finding a large percentage of this was due to cancer. Others suggested the increased mortality rate was most likely due to associated negative lifestyle factors such as lack of exercise or smoking. I think those studies raised more questions than they answered. The pain was associated insofar as it was present, definitely, but causative? I think that's drawing a long bow. You've been around here long enough that I don't feel the need to rehash previous discussions about association versus causation and observational biases, I assume you understand that the presence of a factor doesn't imply causation. For example, if most chronic pain suffers who died in these studies owned goldfish, it would not be reasonable to imply that goldfish ownership increased one's risk of all-cause mortality. Mattopaedia ^{Have a yarn} 12:30, 4 March 2009 (UTC)

I certainly know the difference between correlation and causation (and did long before Wikipedia existed). While correlation alone isn't enough to imply causation, having an explanation of exactly how one event leads to another, and also a proven correlation, together makes a strong case for causation. I've provided several explanations of the mechanisms, and you've now reminded me of another two:

5) Chronic of acute pain may cause the patient to take pain medications, which are likely to cause death if abused, but still may cause death in some cases if taken according to instructions.

6) Chronic of acute pain may cause the patient to "self-medicate" with alcohol abuse, tobacco, or illegal drugs, all of which are associated with decreased life expectancy. StuRat (talk) 15:54, 4 March 2009 (UTC)

Colors of noise images?

Over at the colors of noise article, there are several images of power spectra of the various "colors". While I am not an expert in these I do know the difference between white and pink noise. When I looked at their respective spectra images however, what was called white noise was flat on a logarithmic scale, yet that is the definition of pink noise. I am afraid that several other images may be wrong as well. Could someone with some expertise here check out those images and figure out what needs to be fixed? Thanks, Baccyak4H (Yak!) 17:50, 25 February 2009 (UTC)

I know very little about noise, but there does seem to be a contradiction between the text and the graphs in that article. I think the horizontal scale on the graphs is just labelled incorrectly. --Tango (talk) 18:02, 25 February 2009 (UTC)

It is not just the axis labels; the grid is logarithmic as well. For the "white" noise image, I suppose it could be used to illustrate pink, but again they all look like they were made together, so I worry about all of their accuracy.

I may try and figure out which image is which by reading the descriptions in the text, and shuffling around as needed. It is humorous that the illustration for one "color" might use a file with a different color in its name. Anyway, more eyes, and hands, welcome there. Baccyak4H (Yak!) 18:19, 25 February 2009 (UTC)

I've attempted to sort this out. The correct white noise image was hidden away in the Deleted files archive, so I restored it, and I created a new pink noise image by relabelling the previously misnamed one. The pink noise and white noise articles should now show the correct images.

I think I can see where the confusion arose. The present white noise plot is on a linear frequency scale and the present pink noise plot is on a log frequency scale. It is perfectly possible to show white noise on a log frequency scale and pink noise on a linear frequency scale, if that floats your boat, so there are two ways to show each plot. Someone had got log and linear frequency scales mixed up. --Heron (talk) 19:23, 25 February 2009 (UTC)

Good job. I now reverted colors of noise with the correct images. After checking on the file:image pages of the white and pink images, those are all the pages that use these. Also, as far as I can tell all the rest of the images appear to reflect what their descriptions say. Baccyak4H (Yak!) 19:39, 25 February 2009 (UTC)

These kinds of discussion really belong on the talk page of the article itself. If everyone who had an issue with a science article brought it up here - there would be several thousand new posts each day! SteveBaker (talk) 21:08, 25 February 2009 (UTC)

Are you suggesting we have something better to do with our time? ;) --Tango (talk) 23:50, 25 February 2009 (UTC)

Oh - no! Certainly not. What could be more important than explaining to people that gritting your teeth and ignoring the pain won't save your life in the event of a stabbing-in-the-heart incident? SteveBaker (talk) 01:18, 26 February 2009 (UTC)

(This remark is definitely talk-pagey and shouldn't be here, but I'm posting it anyway.) I see these questions as useful and productive. These questions (and answers, and fixes) show very directly how the rest of Wikipedia benefits from having the Reference Desk. While the first stop for article-related problems should be the article talk page, I see no harm in trying Ref Desk if the talk page approach fails. (I'm prepared to change that stance if we are inundated, but so far the load is manageable.) In this particular case, Talk:Colors of noise has sparse comments about incorrect figures going back to at October 2007, but no one got around to really correcting the problem.

We regularly answer questions of the form, "I read foo on this website/in this paper/on the news, but it seems to contradict bar/itself. What's going on?" If we're willing to do that for every other site on the web, we might as well do it for Wikipedia. At least with Wikipedia, we can go back and directly fix the misinformation afterward. TenOfAllTrades(talk) 15:16, 26 February 2009 (UTC)

Is (or was) Antarctica really a single continent?

Hi. Now, I realise that there really aren't really any strict definitions for a "continent", but let's look at it from a scientific prespective. I know that areas with a single continental shelf is often defined as a continent (but Europe, Asia, and Africa are considered individual continents, and the Americas are split into North and South America), and areas with separate tectonic plates are often considered one continent (as is the case with Asia, India, and Indonesia). However, what I really mean is, do we really know if Antarctica consists of a single tectonic plate, or if there were rifts resulting from a collision of landmasses that stayed together ever since? Old rifts and plate edges could easily have degraded over geologic time, but areas like the New Madrid Fault and the Appalachian Mountains in North America still cause earthquakes today, even though they are both over 100 million years old. One of my hypotheses is that there are far more rifts, faults, and plate boundaries than we know of, even if they don't reach from the surface to the mantle. So, this would make even tectonic plates not have a strict definition. However, could some of these pseudo-boundaries be responsible for some subglacial and other Antarctic features, such as the Transantarctic mountains, Ellsworth Mountains, the mountains east (clockwise) of the Ross Ice Shelf, some Antarctic Volcanoes (I know some are probably caused by hot spots, but hot spots require a weak spot in the crust above the mantle, doesn't they? Also, no tectonic boundaries are known to extend as far south as the continent itself), the Bentley Subglacial Trench, the Gamburtsev Mountain Range? I know that ice can be partly responsible for many great features (such as the Great Lakes), but the ice should act to smooth out the contrasts in elevation, not enhance them. The Gamburtsev Mountains especially seem similar to the Alps in topography, but also potentially in how they formed. Why is this such a mystery? Wasn't that part of East Antarctica supposed to have collided with Australia around the time those mountains were formed, and if not, then could additional plate boundaries be responsible? Could the stress of ice create some new fractures, and if the ice is lifted, could others potentially form (this would imply that global warming could cause earthquakes, because if it can quickly relieve stresses in the crust, why couldn't it cause earthquakes as well)? Do we just assume Antarctica is a single plate with no rift zones or other faults simply because the entire region is covered by a single ice cap that extends to the bottom of the ocean, and because we simply don't have enough information? Thanks. ~AH1^(TCU) 23:06, 25 February 2009 (UTC)

I can't address all of your question but take a look at orogeny which refers to mountains formed by the actions of plate tectonics. There is a section for Antartica. Mountain#geology talks about some of the ways mountains can be formed. This website seems to discuss some of the relevant Antarctic mountain building events. I don't speak geology-ese though so someone else will have to explain the terminology.

I flipped through List of tectonic plates and minor plates which border the Antarctic plate include the Juan Fernandez Plate, the Shetland Plate, and the South Sandwich Plate. The Bellingshausen Plate and Charcot Plate are ancient plates which fused onto the Antarctic plate. The ancient Phoenix Plate subducted under the Antarctic plate and fractured. It seems that the current Antartic plate is indeed a composite of plates. Gondwana seems particularly relevant to Antarctic plate events. 152.16.253.109 (talk) 23:52, 25 February 2009 (UTC)

Also see West Antarctic Rift, which explains those active volcanoes in West Antarctica. If you feel masochistic take a look at The Antarctic Rift: Plume vs. Plate Dynamics. You are right that studying Antarctic geology is difficult due to the covering of ice, the difficulty of travel within the continent, and generally remote location.

With regards to ice, the weight of the ice is significant enough according to a picture caption in Antarctica that removing the ice would cause the landmass to rise by several hundred meters over a few tens of thousands of years. I'm no geologist but that sounds like a fairly rapid rise which would probably cause local faults and earthquakes. I'm not so sure about ice only smoothing out the contrasts in elevation instead of enhancing them. From the article glacier, glaciers deepen mountain valleys they flow through. A lot of glacier related features are fairly steep like fjords, Arêtes, and Glacial horns. It is true that glaciers can't make mountains taller - tectonic and volcanic activity is responsible for that. 152.16.253.109 (talk) 00:24, 26 February 2009 (UTC)