Wikipedia:Reference desk/Archives/Science/2010 February 14

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 14

boiling

when water and milk are boiled, water boils but milk overflows. why?? —Preceding unsigned comment added by 59.96.24.220 (talk) 05:43, 14 February 2010 (UTC)

The milk's cream forms a layer at the top. When the milk boils the steam pushes that layer up making a mess. Dauto (talk) 05:47, 14 February 2010 (UTC)

But is it the increase in the tensile strength of the bubble walls by the protein+fat film on the surface, or the reduction in surface tension? BTW I am not the OP; sorry for hijacking the question this way.--Dr Dima (talk) 05:55, 14 February 2010 (UTC)

It is probably complex. There is probably more than one factor in play. Milk is a complex substance. Probably the bubbles break less easily, thereby they hold their volume, propelling the liquid portion of the substance over the rim of the pot. Bus stop (talk) 06:00, 14 February 2010 (UTC)

... so what is the mechanism for preventing boiling over by placing a ribbed circular metal device at the bottom of the pan? It causes the milk to boil almost like water. How does it work? Dbfirs 07:47, 14 February 2010 (UTC)

All your questions are answered at milk watcher.--Shantavira|^{feed me} 08:17, 14 February 2010 (UTC)

Thanks. I didn't know what they were called. I've never seen the ceramic and glass versions. Dbfirs 08:56, 14 February 2010 (UTC)

As for the "layer" of cream at the top, see scalded milk. ~AH1^(TCU) 03:24, 16 February 2010 (UTC)

Signal White Now - Why?

There is a new product called that makes your teeth instantly white. The manufacturer offer following explanation of its effect: "Signal White Now, our latest innovation, transfers optical-effect technology developed by Unilever’s laundry team to the field of oral care, using a blue pigment to make yellow teeth appear whiter." My question is, why would blue on yellow make your teeth white? (even if it is only for a while). Shouldn't it make the teeth actually green? —Preceding unsigned comment added by Quest09 (talk • contribs) 11:38, 14 February 2010 (UTC)

Yes, it should make them green. Blue+Yellow=White for light, but not for pigments, for pigments it makes green (see additive colour and subtractive colour for details). --Tango (talk) 12:51, 14 February 2010 (UTC)

The OP may be referring to some sort of optical brightener which are commonly used for clothes and do I believe work with the article offering some explaination for how. Or perhaps it is just some simple blue dye as with bluing (fabric) agents such as Mrs. Stewart's Bluing which also work although according to our article have now been largely replaced with optical brighteners. Our article mentions some cosmetic uses for both, particularly for hair. This [1] offers some, not extremely scientific (and obviously not an unbiased source but I don't think there is any real dispute that it works) explaination of why dying white/offwhite blue can make them seem more white. If you overuse the dye, then it's likely to look too blue and if it's something extremely yellow then it's going to look green but this isn't how bluing is done. As an aside bluing agents are also useful for growing rock crystal gardens [2] [3] (also mentioned in the article) although I never found any in Malaysia when I was looking in around 2000 as a suggestion for a friends project. Nil Einne (talk) 14:31, 14 February 2010 (UTC)

Most Americans see something that's truly white as a bit yellowish (I don't have a source, other than my father). My dad helps manufacture plastics, and it's customery to add just a bit of blue coloring to white plastics to make them more attractive. He says that in plastics destined for certain other parts of the world, though, they add different colors. I don't recall what colors get added for what areas, but it's interesting that it appears to be a cultural thing, rather than something hardwired into us. Buddy431 (talk) 17:16, 14 February 2010 (UTC)

This is not exactly how color vision works. The spectrum of light that reaches the eye depends both on the emission properties (such as intensity as a function of wavelength) of the illuminant and the reflection properties (such as reflection coefficient as a function of wavelength) of the surface. The photographic film registers faithfully the light that arrives: use yellow-rich illuminant and the picture will have mostly shades of yellow, use blue-rich illuminant and the picture will have mostly shades of blue. Not so for the human eye. We have a "built-in" color constancy mechanism that allows us to discard (or at least to separate out) the illuminant information and to infer the reflectance of the objects, so that the colors of objects that we see look roughly the same under different illuminants. A gray piece of cloth will look gray both at midday, at sunset, under incandescent lights, and under the fluorescent lights; all the same. The white point therefore is not only subjective, but also depends strongly on the choice of illuminant. Now, yellow and blue are opposite each-other on the color wheel. Thus, adding more blue to the reflection spectrum is indeed making an object look less yellow (more white), as yellowish objects have relatively more yellow and less blue in their reflection spectrum. Oxidation tends to render paper, fabric, and plastics yellow; thus, at least in the past, it was customary to add a small quantity of a blue dye to the laundry rinse water and such. I am not entirely sure, but this is not AFAIK a cultural thing. Color naming is a cultural thing; but color perception is universal among humans with normal trichromatic color vision (which is the vast majority of mankind). And, regarding the question whether you get white or green when you mix blue and yellow, it is explained very well in our Color mixing article. --Dr Dima (talk) 22:42, 14 February 2010 (UTC)

heat loss

a vertical sq. plate 30 cm on a side is maintained at 50 c and exposed to room air at 20 c(temperature) the surface emmisity is 0.8 calculate total heat loss by both sides of plate?

It says at the top of the page, "If your question is homework, show that you have attempted an answer first". This looks like a homework question to me.Heat transfer might help. --ColinFine (talk) 17:22, 14 February 2010 (UTC)

(ec) This looks like a homework question to me, and we don't do people's homework for them. If you have a go at it and get stuck, we'll try and help you out, but you need to show you have tried and tell us precisely where you are getting stuck. --Tango (talk) 17:24, 14 February 2010 (UTC)

Malicious pigeon

The silly title refers to this question: From how high up can one (say: pigeon) drop an object (say: poop) and be certain that it will fall exactly on a given target (say: victim). Let us assume the object is a smooth sphere dropped in still air and we may choose its diameter and weight. Initial velocity is zero. I believe that its trajectory is predictable until either A) due to its increasing velocity the Reynolds number exceeds Re = 0.1 whereupon the path will be deviated unpredictably by turbulence, or B) it reaches terminal velocity while still in laminar flow. I lack an equation for the Reynolds number of a sphere falling in air. I shall assume the pigeon is a perfect marksbird but wonder whether the ultimate bombing accuracy is set by Brownian motion of the surrounding air which would impact accuracy less on a cold day than a hot day. Can anyone help put numbers to these considerations? Cuddlyable3 (talk) 18:26, 14 February 2010 (UTC)

Before you even begin discussing turbulence, brownian motion, and chaos theory, let's start out with an even simpler problem in predicting the actual impact location. Wind can be described (with great simplification) as a vector field for every x,y,z point. At each point <x,y,z>, there is a vector, <wx,wy,wz>(t) - a time-varying windspeed. Assuming you can estimate this 7-degree-of-freedom vector field, you can estimate the force on the falling object at all times. How do you plan to estimate those parameters? Once you have that problem solved, turbulence will be a minor perturbation. The trajectory is predictable only insofar as you know the wind speed at the time that the object passes through all points in its trajectory. I think you will find that people who spend their days laying targets for ballistic objects are satisfied to assume an "exact targeted location, with a point spread". Some will even account for an average wind speed and time-of-flight estimate, to give a first-order correction to impact position. The error caused because you can't estimate wind speed is much larger than any error due to turbulence. If you had, for example, by advanced doppler radar, a reasonably accurate estimate of the wind speed in your volume of interest, you might refine your estimates and need to start worrying about turbulence and nonlaminar flow to increase your accuracy. A better way to phrase your problem would be to ask about the order-of-magnitude error that is due to wind, vs. due to turbulence. On all but the very calmest of days inside the most carefully controlled wind tunnel, I think we could safely say your pigeon would need to worry much more about the former (wind-speed estimation error) than the latter (trajectory estimation error due to non-laminar flow). Nimur (talk) 20:46, 14 February 2010 (UTC)

Nimur your information would have been relevant if I had not said in still air. Cuddlyable3 (talk) 22:58, 14 February 2010 (UTC)

Turbulence happens in still air anytime the object is falling too fast to be in laminar air flow. Turbulence is a chaotic process - hence Nimurs information is highly relevant. SteveBaker (talk) 23:25, 14 February 2010 (UTC)

The Mathematics of Exterior Ballistics, The American Mathematical Monthly (1940), provides an analytic and numerical integral solution based on the Siacci method. The presented model may include any number of parameters, each increasing theoretical accuracy, while practically introducing huge estimation errors. The model includes, in no particular order, gravity, aerodynamic drag that linearly depends on velocity, non-flatness of trajectory, estimation of initial velocity, estimation of air density and air temperature structure, approximation of the projectile as a constant/standard weight, assumption of negligible wind, neglection of the Coriolis force, and finally, error due to numerical integration. Turbulence probably falls into the category of air density and air temperature effects, though it is not mentioned explicitly in this paper. Trajectory Models for Heavy Particles in Atmospheric Turbulence: Comparison with Observations, (2000), Journal of Applied Meteorology, details some more modern methods and mathematics; again, though, "the Langevin model estimated the location and width of the bead deposit swath very well and fixed the peak deposit density to within (at worst) about 100% error" and "uncertainties in the treatment of deposition proved more significant than nuances of the trajectory algorithm." Your insistence on modeling turbulence will categorically put you on a path to a very error-prone subset of all possible models - it would be much more useful to assume turbulence is a statistical perturbation on an otherwise ballistic trajectory, and model the more relevant ballistic effects. In any case, these papers can help "put numbers to these considerations." You might also find the Wikipedia article on external ballistics informative. It presents a large number of common models and some realistic numerical parameters to feed into them. Finally, as you found my earlier comment irrelevant, I'd just follow up and suggest that my response is tailored by experience here - so with all due respect, I may be a better judge of relevance than you're giving me credit for. Your question is asking about a marginal perturbation and neglecting a much more important and realistic effect. It's equivalent to asking the impact of brownian motion on gravity. An effect surely exists - but it's so far below the noise floor, that in anything except a staged, theoretical problem, you would never need to worry about it. Air temperature will have more of an effect on air density and air stratification - which will both affect trajectory - and any effect of brownian motion will be dwarfed by those effects. In any case, today's pigeons are rarely dropping ballistic packages - between precision-guided munition and simple passive aerodynamic munitions, the targeting accuracy is defined by a much more heavily engineered, complicated algorithm than a simple trajectory. Nimur (talk) 03:19, 15 February 2010 (UTC)

Nimur, doubt not that I respect your experience. I reframe my question: a smooth sphere diameter d weight m is dropped from vertical position y=0 with initial velocity u=0 in still air, as would be the case inside a vertical tube (with internal diameter much larger than d). Assume atmospheric pressure, room temperature and Earth gravity. At what (negative) y value is the sphere no longer falling exactly vertically? (Please forget the dynamics of pigeons and munitions and I apologise for introducing the former distraction or implying the latter. This concerns dropping a little ball in laminar flow and not shooting anything.) Cuddlyable3 (talk) 16:58, 15 February 2010 (UTC)

If we assume an ideal sphere with no surface imperfection, and assume an ideal tube with no wall surface imperfections, then the system will be totally symmetric, and any turbulence or nonlaminar flow will be similarly symmetric. The dropped sphere will never deviate from a perfect vertical fall. In reality, the perfect sphere has some nonuniformity - perhaps its mass isn't uniformly distributed, or it has some miniscule scratches or grinding patterns on its surface - and that is what will cause the fluid flow around it to become turbulent asymmetric (which is not the same as turbulent). Your effort to characterize this phenomenon with a Reynolds number can only go so far - remember, that is a quantification of "non-ideal behavior" - but you're trying to set up an ideal problem! So, without some parameterization about the sphere, this is an unsolvable setup. In practice, we would drop the sphere in a test tunnel and measure its properties and construct a table or perform a data fit to some empirical model. In theory, we would apply flow-in-a-pipe equations. But I can't stress enough - when dealing with turbulent flow, your theoretical estimates are going to overwhelmed by estimation-errors that can be orders of magnitude larger than any other parameter you're trying to fit. So, the effort to construct a perfect analytic model of a perfect sphere in perfectly still air is fundamentally flawed. Nimur (talk) 18:01, 15 February 2010 (UTC)

Nimur we need to make more effort to understand one another. I have assumed you understand the empirical significance of Reynolds number. Transition from symmetrical laminar flow to chaotic turbulent flow occurs when inertial forces become too large compared to viscous forces. The onset cannot be prolonged by extreme smoothness of a sphere in fluid which is the simplest case of an object in fluid that has been tested empirically. Equations for fluid Flow-in-Pipe are irrelevant because my setup does not involve movement of air relative to the pipe (which I introduced only to persuade you that the fall is in windstill air). You seem conflicted about turbulence being symmetric; see this. If your response is to be only "Go drop some spheres and do a data fit" then I thank you for that. BTW This question apart from the pigeons relates to a real application and not a geometrical abstraction. Cuddlyable3 (talk) 19:58, 15 February 2010 (UTC)

Are you thinking of reproducing the Book of Tobit? Nyttend (talk) 01:03, 16 February 2010 (UTC)

Tobit 2:10 is about sparrows not pigeons. Cuddlyable3 (talk) 19:48, 16 February 2010 (UTC)

Since Brownian motion etc would be symmetric, Presumably the largest effect causing deviation from a completely straight trajectory would be from the Coriolis effect. Or are we assuming the pigeon takes this into account? 82.132.248.94 (talk) 02:12, 18 February 2010 (UTC)

good alternative to chess?

I love chess, its depth and balance are awesome, but it is very frustrating for me that high-level play implies huge amounts of memorization instead of creativity or mind power, you know what I mean... So does anyone know a comparable game but without that disadvantage? --Belchman (talk) 20:11, 14 February 2010 (UTC)

Go (game)? There have been varients of Go, such as Othello (game), Gomoku, or Pente. Go, and all of the varients, are about as simple to play as Tic-Tac-Toe (or Crosses and Naughts for our Britishy readers) but have very complex and intellectually deep strategies. --Jayron32 20:28, 14 February 2010 (UTC)

[It's "Noughts and Crosses" (or "Naughts..."), for the record, Jayron.] To alter the set-up completely, perhaps Bridge might interest you. Admittedly it's four player, but once you get past a beginner level, it requires the sort of thinking you want (there are several variants which make it almost completely skill and not chance). - Jarry1250 ^{[Humorous? Discuss.]} 20:32, 14 February 2010 (UTC)

Bridge up to the club level, maybe. Beyond that, unfortunately, the contemporary game has become a lot about memorization, as the natural bid has almost disappeared. The first natural bid you make in one of these systems may very well be the contract you intend to play. Defensive signaling is similarly baroque. At an academic level these systems are interesting; in actual play I just find it an annoyance.

I would like to see more Individual events. Without any artificial restrictions on systems. The way I'd run an Individual is, you get to find out who your partner is and talk to him for three minutes. Anything you can agree on in three minutes is kosher, but you have to be able to accurately explain it to the opponents, and if you can't, the risk is on you. --Trovatore (talk) 21:49, 14 February 2010 (UTC)

I agree, the whole "bidding system" stinks of being a way to circumvent the rules - and all manner of meta-rules end up being imposed. I couldn't recommend bridge to someone who wished to avoid memorization. SteveBaker (talk) 23:06, 14 February 2010 (UTC)

The rules do require you to explain your system to your opponents (although usually that's just done by saying "We use the X system with Y variation" or similar). I really don't get the appeal of bridge for the simple reason that I am sure I could write a computer program that would play better than I can, and probably better than most experienced bridge players. Bidding is, as you say, formulaic and the actual play is extremely easy as long as you are capable of memorising the cards that have already been played and all the information you received during the bidding stage. Following algorithms and memorising things are precisely what computers were invented for. --Tango (talk) 23:19, 14 February 2010 (UTC)

Oh no, the play of the hand is not by any means easy. At least finding the best play is not. Granted, on a large percentage of hands, the best play will be found by any player who understands the basic principles and remembers the available information, but that still leaves lots of hands where that isn't true, more than enough to swing a tournament if the bidding is equal.

To me the play of the hand, whether as declarer or as defender, is the soul of the game, and I wish the bidding phase were less important. But it is what it is. The amount of effort put into bidding by the most successful players indicates that that must be where the tournaments are largely to be won or lost. --Trovatore (talk) 00:44, 15 February 2010 (UTC)

I disagree. The first few tricks may have some challenge in them, but after that you have so much information available that it becomes fairly predictable. The problems with the game are evidenced by the prevalence of duplicate bridge in tournaments - there is so little skill in the game (for experienced players) that you can only tell who has the most skill by removing (almost) all the luck. Most games with comparable randomness (eg. poker and Magic: The Gathering, to name two that I have played a significant amount of) you can determine the more skillful players just by playing a reasonable number of games, but that isn't the case for bridge. --Tango (talk) 00:59, 15 February 2010 (UTC)

Well, if you disagree, it means you don't know the game. The play of the has great subtlety and complexity. As I say, it doesn't show up on every hand, but it does on lots of hands.

If you want to find out what I'm talking about, you can't do better than Louis Watson's masterpiece The Play of the Hand at Bridge. You can get it from Amazon UK for about eight pounds. --Trovatore (talk) 01:06, 15 February 2010 (UTC)

I second the vote for "Go". We have computers that can play chess better than the best humans alive - but Go is so difficult that computer Go players are scarcely able to approach the level of professional players. The beauty of it is that there is no memorization of openings - no memorization of end games - and the rules are spectacularly simple. IMHO, it's a much 'purer' game than chess. I think you'll like it. SteveBaker (talk) 23:02, 14 February 2010 (UTC)

Chess computers don't play chess the way people do; the fact that those techniques fail for go doesn't reflect on go's difficulty for human beings. Not that that's an argument against learning go, unless you wanted to learn by playing a computer opponent. -- BenRG (talk) 02:16, 15 February 2010 (UTC)

Or there are variants of chess itself, which are not played at professional level so they don't have that much official and well-researched sequences one has to memorize. Two popular of them in my former high school were:

• White moves once, black moves two times, white moves three times, and so on, always increasing the number by one. Chess is allowed only as the last move in the sequence, and in this case the adversary's first move has to be stopping the chess in one move or being defeated by checkmate. One can win by checkmate or capturing everything except for the enemy king. As the later is much more common, in a tournament the checkmate win can be worth more points than the "simple win". The game is usually over in around 12-13 moves.
• For 4 players, have two chess sets, preferably of similar size. It is played by teams of two, each playing with different colors, side by side. After capturing an enemy piece, you give it to your teammate, who can place it on his/her own board. The rules are:
  • Captured pieces are given to the teammate.
  • When the opponent ended his turn, you can take all the pieces given by your teammate (while your opponent was thinking) in your hand. After this no other captured pieces given by your teammate can be used in this turn, only those already in your hand. (This is to avoid deadlocks which could occur by refusing to make a move until your teammate can capture something you need.)
  • You can choose to place a piece from your hand on the board instead of a normal move. You are not allowed to place something that would put your opponents king instantly in chess, neither can pawns be placed on the last row. Pawns reaching the last row do not promote, they are instantly captured by your opponent. This does not limit the usefulness of the pawns, as a double row of pawns are a very god defense or a good support of an attack, sometimes players ask their teammate to sacrifice a minor piece for a much needed pawn.
  • Achieving checkmate or stalemate on one board instantly ends the game and the situation on the other board becomes irrelevant in scoring.
  • There is a common variant where blocking a chess by placing a captured piece is not allowed.
  • There are other minor variants reducing the chaos in the game, such as only allowing the placement of captured pieces on your own side (first 4 rows) or even more restricting, only on that piece's starting position.

These two game variants are much more chaotic than the original chess, so quick tactical decisions are much more important than lengthy studies and thousands of memorized openings. --131.188.3.21 (talk) 00:39, 15 February 2010 (UTC)

I think you mean "check", not "chess", several times in that post. --Tango (talk) 00:44, 15 February 2010 (UTC)

You are right, I rushed the writing a bit and was not paying enough attention. Not to mention in my native language the two are spelled the same way. I'm sorry, and I hope it's still understandable. I forgot one important rule from the first game variant: "Pawns are only allowed to promote to pieces already captured. In the rare occasion nothing has been captured yet, the pawn is simply lost." And i hope it was clear enough that the second, team based game is completely different, with normal "one move per turn" rules. Enjoy! :) --131.188.3.21 (talk) 00:50, 15 February 2010 (UTC)

There's also Chess960 where the opening positions of the pieces are randomised. This makes it very difficult to memorise openings. - Akamad (talk) 01:49, 15 February 2010 (UTC)

In fact, Chess 960 was developed for that very reason. It's not as fun as ordinary Chess though. There are many other Chess variants to choose from. Zain Ebrahim (talk) 08:10, 15 February 2010 (UTC)

Cribbage and a special favourite of mine, Stratego. --Dweller (talk) 10:46, 15 February 2010 (UTC)

For a break I often play online Reversi, as it is a quick game that only lasts a few minutes. If you have Windows, then you may already have it installed under games, or you may be able to download it from Zone dot com. I wish it was possible to choose different board sizes and shapes to make the game more varied. 89.240.201.172 (talk) 13:59, 15 February 2010 (UTC)

Kasparov advocates Advanced Chess, in which one human/computer team competes with another human/computer team.

Frankly, I'm not convinced. I think Kasparov overstates the role of "Creativity" and the human element in Chess. Especially after his allegations that Deep Blue was playing "creatively". (And therefore cheating.)

But what do I know? I'm certainly not even close to being a Grand Master. APL (talk) 16:11, 15 February 2010 (UTC)

Thanks for your replies. I have been considering Go as a possible alternative for a long time, but the problem is that I prefer the more tactically-oriented or (sometimes) aggressive nature of chess to the strategically and slow-paced nature of Go. This is a serious dilemma that I have. --Belchman (talk) 16:25, 16 February 2010 (UTC)

why is formic acid so much more toxic than acetic acid?

I don't really get it ... does formic acid have a greater affinity for dissolving into biological tissues? John Riemann Soong (talk) 21:23, 14 February 2010 (UTC)

I believe the pH level of formic acid is 2 (more toxic), and acetic acid is 3. -^Avicenna_sis @ 21:38, 14 February 2010 (UTC)

I went to google, typed in "formic acid" toxicity and on the very first page of hits was an except of an article "Formic acid (or formate) is apparently more toxic than other fatty acids, possibly owing to its enzyme-inhibiting activity". The wikipedia article formic acid states that it specifically affects a certain biological structure--that would be another useful search-term to help you find your information. DMacks (talk) 21:44, 14 February 2010 (UTC)

Miller-Urey

Miller-Urey experiments with better predictions of Earth's early atmosphere produce most (not all) amino acids and other organic compounds. However, they also produce toxins like formaldehyde and hydrogen cyanide which would kill any life which did happen to form. How are these experiments considered evidence for abiogenesis/evolution when they fail to produce all of the necessary compounds, and they also produce horrible poisons? --70.129.187.17 (talk) 21:59, 14 February 2010 (UTC)

All "necessary compounds" for what purpose? Remember, life doesn't have to be something we recognize to match our current life-forms in terms of their complete biochemical makeup. Likewise "toxins" is only in your current experience of certain biochemical processes. See for example, Hydrogen cyanide#HCN and the origin of life. DMacks (talk) 22:23, 14 February 2010 (UTC)

What is toxic to you may not be toxic to early life. There are plenty of extremophiles that live in places inhospitable to human yet it is still life. Also something to note is that evolution has nothing to do with abiogenesis. Evolution describes a continuous adaptation of life to its environments; it does not nor does it attempt to explain the origin of life itself (Darwin's book is called "On the Origin of Species" for a reason). --antilived^{T | C | G} 22:30, 14 February 2010 (UTC)

(ec) The primitive life forms which would have hypothetically arrived during Miller-Urey type conditions would be far too basic for HCN and formaldehyde to be "toxic" to them. I'm guessing they would just be a membrane with some nucleotides inside. HCN is toxic to us due to interactions with mitochondrial enzymes, I believe. It also should be noted that there is a hypothesis for the role of polycyclic aromatic hydrocarbons (which are carcinogenic) in the origin of life (see PAH world). This is another example which shows how fallacious it is to assume that "chemicals toxic to current organisms = chemicals toxic to primitive organisms". --Mark PEA (talk) 22:34, 14 February 2010 (UTC)

Firstly, (as our article points out) subsequent recreations of the Miller-Urey experiment produced all 22 standard amino acids. Secondly, the experiment is supposed to prove that it's possible for these compounds to have formed spontaneously from a plausible set of initial conditions. It is emphatically NOT claimed that "this is how it happened". There are a trillion possible variations of temperature, pressure, chemical precursors and energy sources that could have produced the actual abiogenesis event - and it's unlikely that we'll ever know precisely what happened because conditions changed repeatedly and dramatically over the billion years between the formation of the Earth and the first fossilised lifeforms that we have discovered.

This experiment should be viewed as a "proof of plausibility" rather than a "recreation". In dozens of other experiments with wildly different starting conditions, we get very similar results, showing that even if the conditions were quite different from the Miller-Urey experiment, there can be no doubt that these compounds would have been present in early earth history.

As for the poisons - we know that life can survive and evolve in an amazing variety of conditions (see Extremophiles) and it's no stretch at all to imagine that it could have formed despite those seemingly nasty compounds.

• This site suggests that cyanide was actually part of the chemical pathway to make Purine Adenine.
• This PBS documentary described how extremophiles survive in cave with normally lethal doses of formaldehyde.

So it's perfectly plausible that life could have formed even in such a seemingly-toxic environment. Also, we do not know what other reactions or physical separations were going on that could have subsequently eliminated these other compounds. Since we know that life most definitely exists and that it certainly evolves, it's really only necessary to find a plausible path to that initial abiogenesis event to complete the story to an acceptable degree of scientific proof. SteveBaker (talk) 22:56, 14 February 2010 (UTC)

Cyanide would actually have been pretty useful if you were trying to form the first cells. It's a basic building block for organic substances. Do you know why cyanide is poisonous to us? Today we're obligate aerobes and we need aerobic respiration. A long long time ago, oxygen was a poison whereas normally poisonous gases (to us) like methane, phosgene, etc. would have been useful. John Riemann Soong (talk) 21:57, 15 February 2010 (UTC)