Wikipedia:Reference desk/Archives/Science/2016 November 26

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.

November 26

Feedback (positive) !!

Guys, believe it or not, I had never visited this page before. The first time I checked the page out was yesterday. I frequent the Help desk considerably, helping out in resolving queries and all that stuff. The general rule out there is that if it doesn't relate to Wikipedia, nuke it. So it was a pleasant surprise seeing the queries here, and even better, seeing the responses, from intergalactic stars to pressure cooker issues. Very interesting, informative and in some cases, absolutely rollicking answers. Just thought I'll leave a note of appreciation. Thanks for making it so interesting. Lourdes 01:49, 26 November 2016 (UTC)

Comments like this really belong on the talk page, and that page serves to cover the entire ref desk, not just the science one. We'll forgive you this one time, given how great your song "Royals" is. :) μηδείς (talk) 04:27, 26 November 2016 (UTC)

On behalf of all of us, thank you Loudres for your kind comments! Dr Dima (talk) 04:46, 26 November 2016 (UTC)

This being the Ref Desk, I can't help but add a link to positive feedback. :-) StuRat (talk) 17:29, 28 November 2016 (UTC)

puzzled about theta Eridani

Its two components are designated θ¹ Eridani, also named Acamar, and θ² Eridani.

This is the first time I've read of stars with a digit attached to their Greek letter that are not distinct to the naked eye, or of a binary in which a traditional name is attached to one rather than both; usually binaries are distinguished with A/B after the collective name. Have I misunderstood the system? Is θ Eri a special case? —Tamfang (talk) 07:22, 26 November 2016 (UTC)

Were they further apart in the past? Sagittarian Milky Way (talk) 07:53, 26 November 2016 (UTC)

They're now 8+ arcsec; how close can a naked-eye double be? —Tamfang (talk) 00:21, 28 November 2016 (UTC)

Depends on your eyesight but 8 seconds of arc isn't it. The world record is 20/5 vision (previously thought to be sub-threshold) which would be about 15 arcseconds. Sagittarian Milky Way (talk) 03:03, 28 November 2016 (UTC)

Wikipedia says at Bayer designation#Other Bayer designations that:

A further complication is the use of numeric superscripts to distinguish neighboring stars that Bayer (or a later astronomer) labeled with a common letter. Usually these are double stars (mostly optical doubles rather than true binary stars), but there are some exceptions such as the chain of stars π¹, π², π³, π⁴, π⁵ and π⁶ Orionis.

This is unsourced, but seems plausible. Perhaps someone assumed that Theta Eridani was an optical double. Or perhaps the A/B convention and the 1/2 convention competed for a while and were never made systematic. This article tells something about how chaotic the history of star designations has been. --76.71.5.45 (talk) 13:22, 26 November 2016 (UTC)

θ Eri seems to have acquired its superscripts rather recently; they're not in Allen's Star-Names of 1899. —Tamfang (talk) 23:58, 27 November 2016 (UTC)

The definition of amino acid

According to what I understand the difference between peptide and peptide bond is that peptide is two or more amino acids (up to 19 -inc.) which linked together, while the term "peptide bond" refers to the bond between each two amino acids. Is that true? If it is, the definition of protein is that it contains at least 20 amino acids which are linked together. Isn't it? 93.126.88.30 (talk) 11:01, 26 November 2016 (UTC)

Peptides are biologically occurring chains of amino acid monomers linked by peptide (amide) bonds. An oligopeptide consists of only a few amino acids (between two and twenty). (Poly)peptides that contain less than 20–30 residues are rarely considered to be proteins. Proteins are generally much longer Polyamide chains ranging in size from tens to several thousand amino acids (in a sequence built from 20 standard amino acids) folded into 3-dimensional structures. See the article Amino acid for more information. Blooteuth (talk) 14:45, 26 November 2016 (UTC)

One small quibble. There are several amino acids that are capable of forming three peptide bonds, so in principle it might be possible to have polypeptides that are tree-shaped rather than linear. These would not be considered proteins: a protein is a linear chain of amino acids linked by peptide bonds. Looie496 (talk) 15:01, 26 November 2016 (UTC)

After reading the entry of "peptide" now I'm more confused - because I saw so far so many opinion the definition. How could it be that there is not one clear definition for peptide that differs it from the protein? "Peptides are distinguished from proteins on the basis of size, and as an arbitrary benchmark can be understood to contain approximately 50 or fewer amino acids.[1][2] Proteins consist of one or more polypeptides arranged in a biologically functional way, often bound to ligands such as coenzymes and cofactors, or to another protein or other macromolecule (DNA, RNA, etc.), or to complex macromolecular assemblies.[3] Finally, while aspects of the lab techniques applied to peptides versus polypeptides and proteins differ (e.g., the specifics of electrophoresis, chromatography, etc.), the size boundaries that distinguish peptides from polypeptides and proteins are not absolute: long peptides such as amyloid beta have been referred to as proteins, and smaller proteins like insulin have been considered peptides."93.126.88.30 (talk) 04:20, 27 November 2016 (UTC)

In many fields of science, what seems to outsiders like a fundamental problem of such a basic definition in the field actually isn't important at all. For example, suppose one could define what exactly a peptide is as opposed to a protein. That wouldn't really change the field itself because that level of specificity isn't of interest when those terms are used. Instead, we often focus on the specific length (or range) as a number, or the range of weights where the exact bonding details aren't relevant for separation, or a specific set of polyamides where the relevance to protein (as a function) isn't relevant. Proteins generally are larger (but there is indeed no firm cutoff solely due to length) and may contain more than one chain, and structures other than peptide chains, and have bonding other than peptide bonds among the amino acids. Lots of terminology has just grown over time, and coming from different perspectives, such that eventually it's not consistent and unless there's a scientific problem caused by it...meh. As additional examples, nobody can agree (and most agree that nobody actually cares, and everyone knows that the historical separation is completely bogus) what makes something an Organic compound vs inorganic. DMacks (talk) 04:35, 27 November 2016 (UTC)

The peptide bond refers to the actual chemical bond between carbon and nitrogen that is formed as a peptide is made. The hydrolysis of a peptide, for example, might be described as breaking a peptide bond, even as it produces two smaller peptides (or a peptide and an amino acid, or in the case of a dipeptide, two amino acids). It is basically a way to 'navigate' while describing the peptide as a molecule. The peptide simply is the molecule, the whole molecule. So if you look at the R chain for any given amino acid, you might say it is not on the peptide bond, but it is on the peptide - it is attached to the carbon adjacent to the peptide bond (i.e. the alpha carbon). Wnt (talk) 14:05, 27 November 2016 (UTC)

Could a tectonic tsunami from the tropical Pacific enter the Atlantic?

I think not, the lowest routes would be in Nicaragua or Panama and those are at least 50 miles long and tens of yards high. Am I right? There are no fjords I know between the high Andes and Sierras to funnel tsunamis into a 1/3rd mile high flood. Sagittarian Milky Way (talk) 16:34, 26 November 2016 (UTC)

The Lituya Bay event was that high over a very short distance and caused by a massive landslide into a narrow inlet (albeit triggered by an earthquake). The biggest recorded tsunami that I know of was the 2004 Indian Ocean earthquake and tsunami, with a maximum of about 50 m. The lowest point on the continental divide is the Riva isthmus (about 50 m elevation and the potential location for the western section of the Nicaraguan Canal), which could theoretically be crossed by the largest of tsunamis, but that only takes you to Lake Nicaragua - a long way away from the Atlantic. Tsunamis run out of energy pretty quickly going uphill, which is why it's possible to run away from them with sufficient warning in many cases - not so in Banda Aceh of course, which is very flat. Mikenorton (talk) 22:23, 26 November 2016 (UTC)

About how small of an asteroid could do it? Maybe an oblique trajectory pointing towards the coast? Sagittarian Milky Way (talk) 22:42, 26 November 2016 (UTC)

Don't forget about passing S of S America or N of N America. I suspect that all tsunamis do so, to a very minor extent. StuRat (talk) 18:19, 27 November 2016 (UTC)

Whether a tsunami can pass north of North America depends on the season and whether the ocean is sufficiently ice-free to permit passage of the tsunami. The Drake Passage does not freeze over, and I would expect that a tsunami could pass south of South America. Mariners who sail the Drake Passage, or the extreme southern waters in general (but there is no need to sail that far south except to round Cape Horn) refer to a phenomenon known as the thousandth wave, which is apparently an occasional rogue wave. Given that the so-called Southern Ocean goes all the way around between Cape Horn and Antarctica, once a rogue wave starts circulating from west to east, it may just continue going around the world. Robert McClenon (talk) 03:40, 28 November 2016 (UTC)

See Antarctic Circumpolar Current. Robert McClenon (talk) 03:44, 28 November 2016 (UTC)

• The NOAA animation at 2011 Tōhoku earthquake and tsunami shows the wave rounding Cape Horn. μηδείς (talk) 23:57, 29 November 2016 (UTC)

Can two neurons (or networks) have infinite possible connections?

A connection between two neurons could either not exists, or be one point between 0 and the strongest possible connection? They could also be at any distance from each other (given a range of distances between 0 and x). There are infinite values in any range. --Hofhof (talk) 19:48, 26 November 2016 (UTC)

A direct connection between two neurons is either chemical (a chemical synapse, or "synapse" for short) or electrical (a gap junction). While a maximal signal in either case is limited by the properties of the two cells connected, the number of possibilities is mathematically infinite because you can, if you so desire, define the strength of the connection with an arbitrary high precision. In other words, the number of possibilities is, indeed, as infinite as the number of real numbers between 0 and 1. However, this does not mean that this high precision is either meaningful or necessary. Indeed, operation of a chemical synapse relies on release of neurotransmitter, which is stored in vesicles in the presynaptic terminal. With each activation of the presynaptic terminal, only a few vesicles are released (or sometimes none at all), in a process that is currently thought to be at least partially stochastic. In other words, the chemical synapses are neither perfectly predictable nor perfectly precise. Therefore, if you build a model of a certain neural network, and the model requires you to tune the connection parameters to an unreasonably high accuracy to work properly, then it is likely that you are doing something wrong or at least non-biological. Biological neural networks tune themselves to the right ballpark of parameters by a variety of homeostatic mechanisms, but most networks are robust to perturbations such as a death of an individual neuron. Model networks should also be robust to stochastic imprecision of the synapses, so you may want to include this stochasticity into your model. Dr Dima (talk) 20:12, 26 November 2016 (UTC)

The "infinity" of values there is probably exaggerated, since the accuracy of biological numbers is not very great. (i.e. if you turn up the subwoofer, the distance your axon travels gets longer and shorter as the low frequency sound waves penetrate the relevant structures. Changes in blood sugar level affect how readily signals are sent, while ionic strength influences ion flow etc.)

A more meaningful range of variation is obtained because neurons can attach to one another in different ways. If one synapse is further out on the dendrite than another, the second synapse might be inhibitory and prevent its signal from getting through. Or the activating signals from the two might sum up - especially if the further one is received a little before the nearer - and put the cell over the potential for an action potential. Or... well, there are a lot of things, and it's been a very long time since I reviewed this topic. But you should look that way. Neurons are covered in synapses; it's an unbelievably complex network. Things like the Cajal stain (hmmm, Golgi stain) work by randomly showing you a tiny fraction of the neurons present because if you could see them all stained the tissue would be just black. Wnt (talk) 13:55, 27 November 2016 (UTC)

In AI in computing they generally find a byte of 8 bits is quite sufficient for weights though 16-bit floating point values are also common. Great accuracy is not needed and I doubt the bags of mostly water that are our cells depend much on accuracy. Dmcq (talk) 14:20, 27 November 2016 (UTC)

The lees liquid in the body - the more sugar concentration and decreased freezing point

I saw a citation from a prof. who said the next things (translated): "when you have cold, one of the first things that your body does is to cause to you to get rid off the liquids, such us in urination. This is a simple equation: the less liquids you have, the percentage of the sugar increased and the freezing point decreased.". Well, I don't understand 2 things that he mentioned: 1) why or how the percentage of the sugar depends on the liquids of the body? 2) what is this "freezing point" that decreases?(does it mean that the less water the smaller freezing point"? if it does, it should not be opposite? it's more difficult for a river to be frozen than a drop on the ground.) 93.126.88.30 (talk) 21:00, 26 November 2016 (UTC)

Read our articles Freezing-point depression and Fluid balance. Graeme Bartlett (talk) 21:31, 26 November 2016 (UTC)

I am skeptical of this explanation. It is true that eliminating water and making the serum more concentrated will decrease its melting point. However, looking at two totally different formulas for the expected serum osmolality [1] [2], what they have in common is that the concentrations of sodium, glucose, urea and in one potassium are used to come up with a figure that, in the second, is about 280-300 mOsm, (this says 275 to 290 - if I were anywhere near serious I'd need better sources than any of these, but this is strictly back of the envelope/order of magnitude) where apparently 1 Osm = 1.86 K freezing point reduction (that number being the cryoscopic constant). So we're talking about not much more than a degree difference between normal serum and pure spring water. You can't live with pure spring water in your veins, but I'm not sure how much stuff you could tolerate having on the other side - but hypernatremia (and hyperkalemia) is not good, and the sodium apparently accounts for most of the figure. That last link says that under 145 mmol/l is normal and over 160 mmol/l puts you in the emergency room, so that leaves you with basically 10% room for play in manipulating the average serum water content to avoid having your blood freeze, which amounts to not much more than .1 degree Fahrenheit if I understand this correctly. There would seem to be more merit of having the body work as it should to support healthy shivering and sheltering so its tissues aren't freezing! Wnt (talk) 00:38, 27 November 2016 (UTC)

segway design useful for hovercraft?

Hello, I'm wondering if the way the segway was designed with gyros, microprocessors?, sensors and electric motors could be or has already been used to improve performance and reduce cost of hovercraft--for one thing, my understanding is that sufficient height and sufficient stability compete with each other right now. Thanks.108.252.141.219 (talk) 23:54, 26 November 2016 (UTC)

The Segway applies an active feedback control system to stabilise its mechanical system of an Inverted pendulum consisting of the rider's Center of mass supported above the wheel axle, with inherent instability in his rotation on that axis. Hovercraft do not have this instability since their center of mass is supported above a large-area cushion of pressurized air. In still air on a level surface, the hovercraft location is stable but it can be moved horizontally by a very small force, almost exclusively that given by Newton's 2nd law: $F=$ mass x acceleration. In windy conditions or on sloping ground, keeping the hovercraft stationary requires compensating force e.g. by ducted thrust fan(s), that may be controlled manually or by an automatic Dynamic positioning system whose components may include motion sensors, Gyrocompass, error computer and Servomotor. A fundamental tradeoff in hovercraft design concerns maintaining an unbroken air cushion: the total amount of air needed to lift the craft is a function of the roughness of the surface it travels over. With adequate air cushion height, ground irregularities that displease a segway traveler are imperceptible to a hovercraft passenger. Many hovercraft designs were completed in the 1960s while the Segway patent (US Pat. 6 302 230) was granted in 2001. Blooteuth (talk) 16:56, 27 November 2016 (UTC)

Some VTOL aircraft are examples of an unstable system which needs active feedback controls. StuRat (talk) 18:23, 27 November 2016 (UTC)

Yes. But that's rocket science. Blooteuth (talk) 18:58, 27 November 2016 (UTC)

Not necessarily -- ordinary helicopters (especially the single-rotor type) are unstable in flight and need constant active feedback from the pilot in order to remain under control. 2601:646:8E01:7E0B:4C25:8F4F:2BC7:C702 (talk) 01:22, 29 November 2016 (UTC)