Salt (sodium chloride). The uses and misuses of salt in freshwater aquaria rate their own pages.
In the fourth century BC Aristotle placed freshwater fish into seawater to observe their reactions. The aquarist's oldest treatment for freshwater skin and gill parasites is salt. "Salt is Nature's remedy for many ills," wrote William T. Innes in Exotic Aquarium Fishes, 1935, and salt as a medication survives from that era, when aquarists had very little else to turn to.
A salt dip or a salt bath still has uses. Careful! an effective salt bath will kill many plants. Conversely, how effective on ciliate parasites is a salt bath that doesn't kill plants? In planted tanks like mine, an effective salt treatment would be ruinous: a potassium permanganate treatment is my alternative.
Temperature should be raised to 80oF, or higher as long as the high temperature isn't stressful itself to the particular fish in question. Mr Innes' recommended concentration began with two level teaspoons of salt per gallon (the equivalent of 0.2% salinity) and built up gradually over 24 hours to four teaspoons/gallon. If no improvement was noted, by the third day you could go to six (that's two level tablespoons), as long as the fish were showing no signs of distress. "At the end of treatment," Mr Innes cautioned, "slowly add fresh water until the salt content is low before returning fish to the aquarium."
Modern variants of the salt dose vary: a common version such as that at Doc Johnson's KoiVet recommends just half of Mr Innes' highest concentrations.
Salt tolerance in freshwater fishes varies. The percomorph fishes, like cichlids and anabantoids, are derived from marine ancestors in the age of dinosaurs. In general, they are more salt-tolerant than ostariophysii, the loaches and minnows, characins and catfish that have descended from distant freshwater ancestors. Years ago, William T. Innes reported that when a range of freshwater fishes were exposed to salt baths, the first to die were Corydoras. Those ostariophysan fishes that navigate by electric fields, like knife fishes, gymnotid eels and the "elephant noses" or mormyrids, should never be exposed to salt baths, according to the University of Florida. Killifishes and livebearers are more tolerant of salt, on the whole; small tetras are less tolerant. But there are many exceptions to this very broad rule. Be aware of the salt tolerance of your particular fish.
Salinity measurements. A 3ppT or 5 ppT (parts per thousand) salt solution is equivalent to 0.3% or 0.5% salinity, or 3 or 5 grams per liter, or 3 or 5 teaspoons of salt per gallon. This is the standard recommendation for salt treatment for skin parasites and for Ich that you will find at Doc Johnson's KoiVet. Remember, a level tablespoon is equivalent to three level teaspoons. 
Salt basics from U. of Florida. If you truly want to consider the effects of salt, it's good to know where to look. So start off by reading Dr Ruth Francis-Floyd's article "The use of salt in aquaculture," the basic information presented by the U. of Florida's Institute of Food and Agriculture Science for the use of Florida's many fish farmers. The article outlines benefits of rock salt (a.k.a. kosher salt), both in controlling some parasites and in minimizing nitrite stress during shipping and handling.
Robert T. Ricketts' article, "The Salt of the earth... the salt from the sea," from Tom Griffin's former AquaMag, now archived at the AquariumWiki, is essential reading. Exploding Urban Myths about toxic additives and salt as a tonic as he goes, RTR also gives good explanations of general hardness (GH) and alkalinity or "carbonate" hardness ("KH").
Salt as a "preventative". Some aquarists have always maintained a low level of regular table salt, NaCl, in the aquarium water, to "prevent" disease. Typical levels range from one teaspoon per ten gallons to as much as one tablespoon per five gallons. If you have any doubt about whether this is a magical practice or not, just try to dissuade someone from adding that pinch of salt, "just in case." From the reaction you get, you'll quickly recognize that you're in the land of magic. RTR made some useful points about rock salt, compared to marine salts in an AC post, Nov. 2000: "Brackish and estuarine fish are able to tolerate and likely to prefer more than only slightly higher Total Dissolved Salts, or osmolarity. In nature this is coupled with increased pH from the other salts dissolved in the water: calcium and magnesium carbonates and bicarbonates prominently. These materials "buffer" the water at higher pH levels than waters lacking these ions. So in the wild these fish arose in or frequented waters with ALL these ions and more. Adding salt, NaCl, alone will not affect the pH at all. Adding marine mix will provide a situation more like that from which the fish's forebears came. So long as the levels are below those that would register on a full range hygrometer, short-term exposure will do little harm (disease treatment), but long-term exposure should be reserved for fish with the physiology to live in that situation. I suspect that part of the reason many livebearers do well in this situation is that it is also more stable than an unbuffered tank, and for many fish, stability is as important as absolute pH, GH, or KH, provided it is within a ballpark of their "natural range". While some salt will certainly not peel off a cory's armor, it is not right for the fishes' physiology. Short-term is okay (provided you don't make sudden changes of too large a magnitude), long-term is likely to mean shorter life — you can call it another unwarranted stress. And by the same token, water without the full mix can be a stress for those adapted by physiology over generations to having such conditions."
"I'm far from a purist about water," RTR continued, "and rarely suggest folk go through the hassle of doing water modifications. I seldom do so myself. Most fish are quite adaptable. But if I do water mods, I want them to be in the right direction and for sound and understood reasons. I will not do so because popular myth says it is good for all fish in all water at all times. I know better than that, and wish more hobbyists would think about why they do the things they do, especially when they are adding things to the water."
Here's a useful warning from Nomad: "I have treated fish for Ich using the salt/heat method, bringing salt concentrations up to ~1 tblsp/5 gal concentration. The trick when doing this with corys in the tank is to go slowly. Predissolve the salt in tankwater, and add it back into the tank a little at a time over several hours. When I've done this, I've never seen any signs that the fish in question were in any stress whatsoever. Just dumping in the salt, OTOH, can lead to them going belly-up pretty quickly. As with other parameters (pH, hardness, etc.) the real stress is not generally caused by the actual values in question, but by the speed of the change. Fish should be acclimated to higher salt levels the same way they're acclimated to differences in pH or hardness."
A note about salt and evaporation. You know that salt doesn't evaporate from aquarium water, of course. Though you may ignore my advice, and add a little salt to the aquarium, you know better than to re-salt the water after each partial water change, of course. So you are careful to add it only to the make-up water each time, at the recommended dosage. But! you need to top up the tank first, with unsalted water, to make up for evaporation, or else you are replacing evaporated water with salted water at each partial water change. The additional dose of salt depends on the percentage of water exchanged each time.
Salt dips and salt baths. Solutions of salt — sometimes right up to marine concentrations (35 parts per thousand) — are recommended as a short-term prophylactic bath for half an hour to remove parasites from gills, from fins and the outer surface of the epidermis, as Mr Innes recommended so long ago. Encysted parasites, like mature Ich, are less likely to be affected. Intestinal parasites aren't affected at all. Chicago's Shedd Aquarium bathes their new freshwater arrivals in full-strength seawater to rid them of ectoparasites, without losses, said a curator there in a TFH interview a decade ago. It's smart to pass your own new arrivals through a half-hour salt bath before putting them in the quarantine tank. A five-gallon covered bucket with an aerator is what you need.
The salt concentration has to be strong enough to make gill flukes drop away. Build up your dosage with pre-dissolved Kosher salt in two or three increments and stand ready to do a 50% water change with aquarium water at the first signs of stress; heavy respiration is an early signal. You'll want to do a 50% water change anyway, when the time's up, before netting them into the QTank.
Rock salt. Kosher salt. "Aquarium" salt. Marine salt. Rift Lake salts... What kind of salt to be using? "Salt" is the generic term for the dissolved ions and mineral component of all natural waters, which covers salts like Epsom salts, which is magnesium sulfate, characteristic of the spa waters at Epsom in England. (No, not "Epsom's" salt; "Dr. Epsom" will not be in today — or ever!) "Sea water is still better than salt crystals," Mr Innes wrote in 1935. I'd say "different" rather than "better." Sodium chloride represents only about 77% of the total dissolved solids in seawater, and marine salt's dissolved carbonates will raise your pH. When we're talking about just plain "salt" we mean NaCl, sodium chloride, which is kosher salt or rock salt. "Animal feed grade" rock salt meets high standards of purity. "Halite" is simply a lower grade of rock salt. Impurities may color it gray or even brownish. The "halite" sold for de-icing driveways, etc. isn't pure enough for aquarium uses, as you already surmised. "Table salt" often has iodine added, not the mythic poison it's made out to be, an issue separately addressed in a few moments. Table salt often contains as well some free-running additives, like magnesium carbonate, magnesium stereate or silicon dioxide — all harmless. "Aquarium" salt is nothing but marine salt ("sea salt") that has been repackaged for the hobby and presented by colorful non-union toons in lab coats; check the label. Marine salts and Rift Lake salts contain many trace elements ("electrolytes") in addition to plain therapeutic NaCl. They are not substitutes for rock salt, nor vice versa. Nor substitutes for one another. Rift Lake waters actually contain surprisingly low concentrations of sodium chloride. As RTR pointed out, the other solutes in marine salt and Rift Lake salts will boost pH and hardness. You don't want to needlessly change those parameters as part of a medicating salt regimen.
Now I even hear of "freshwater salt!" Cheeky! But don't completely ignore the salt packaged for the aquarium hobby: "Aquarium" salt also makes great pasta! And when guests drop over, set the "Aquarium" salt box on the table next to the pepper mill for a sparky conversation starter! They'll be impressed that you paid so much for salt!
Iodized salt. RavenSedai posted at Tom's Place, 23 July 2000, the following from a koikeeper newsgroup, which pinpoints the anxiety about iodine in salt, an anxiety I don't share myself. Here it is, so you can decide whether to avoid iodized salt or not: "Iodized salt has 77 ppm iodide for the prevention of goiter. Ocean water typically contains 0.006 ppm iodide, so sea salt is typically 0.2 ppm iodide. So if iodized salt were used in the pond, the iodide concentration would be roughly 400 times higher than if sea salt were used to produce the same salinity. In natural waters (ocean or freshwater), the predominant form of iodine is the oxidized form, iodate (IO3-), not the reduced form, iodide (I-). These two forms do not interconvert readily in water. The active ingredient in Betadine disinfectant is 1% iodide. The USDA suggests a 10 minute bath of 100 ppm iodide as a disinfectant for trout eggs. A permanent concentration in the pond of about 0.1 ppm, a weak disinfectant bath, sounds like an experiment, not like prudent advice. All of the USDA aquaculture publications specifically recommend non-iodized salt. The toxicity of iodide to fish is unknown. It seems likely that fry might be more affected than mature fish. In humans, toxicity begins at about 2 mg/day, only 13 times the US FDA recommended daily allowance. For that reason, salt is iodized at much lower concentrations in Europe than it is in the US. Finally, why suggest experimenting by using much more expensive iodized table salt? I bought an 80 pound bag of solar salt crystals at Lowe's for $2.87. It's much cheaper, and safer."
Fears about toxicity of the iodine represented in table salt are still often expressed in warnings not to use iodized salt in the aquarium.
Potassium iodide (sometimes it's sodium iodide) in U.S. iodized table salt ranges from 20 to 40 parts per million. So what part per million of iodide does that potassium iodide represent? I'm innumerative. The ppm iodide were worked out by aquariaddictus in a thread at AquariaCentral, started 1 March 2003 (you can find it at AC: search "iodized salt ppm iodide"). Aquariaddictus pointed out that there is no elemental iodine in KI, just as there is no free chlorine in table salt. Iodide is I-, while iodine is I2, as chloride is Cl- while chlorine is Cl2. The terms aren't interchangable. Potassium is number 19 on the periodic table, Iodine is number 53. So pure KI is 73% iodide. Thus the iodide in the potassium iodide additive is between 14.6 - 29.2 ppm in the dry salt. KI weighs 15.38 grams per teaspoon. So, 15.38 grams/one teaspoon per gallon = 15.38 grams per 3.7854 liters = 4 ppm as KI. If you add iodized salt at the rate of one teaspoon per gallon, you are adding iodide in the range of 0.083 - 0.166ppm. After laying out the calculations, aquariaddictus remarked, "All in all, I have to believe it's a drop in the bucket. Does anyone use a tablespoon/gallon except in times of severe disease?"
A salt resource. The Salt Institute, a non-profit association of salt producers, maintains a website with lots of authoritative information about sodium chloride.  
Some genuine effects of salt in freshwater: osmoregulation. If salt has an effect on freshwater fish, how would that be produced? Well, salt dissociates in water to its two ions, sodium (Na+) and chloride (Cl-). Normally when you're talking about beneficial "effects" of salt in aquarium water, you're talking about the effects of the chloride ion. A fundamental effect of salt is upon osmotic regulation. Osmoregulation, as it's called, is the interior balance of ions ("electrolytes"), the dissolved salts in blood and tissue that each living organism must maintain. "Osmoregulation" is the regulation of osmosis, that is to say, the selective diffusion of ions through a semi-permeable membrane. The gill surface is an example of a semi-permeable membrane. So is the coating of a bacterium. Osmoregulation is as important on the cellular level as it is to a multicellular organism. Salt concentration in the water affects the cells' ability both to retain water and solutes or to expell water or solutes: a more dilute solution — freshwater, for example — tends to diffuse across a membrane into a more concentrated solution, such as fish blood. The tendency is to equalize concentrations of solutes. Individual cells — and fish eggs and sperm are single cells — are less tolerant of osmotic pressure than multicellular organisms, most of whose cells are constantly bathed in blood and lymph fluids, which protect the cells with an osmotic and pH-stable environment. We take advantage of this vulnerability of single cells when we consciously stress a unicellular parasite with salt. And we are sometimes disappointed by the low hatching rate of tetra eggs in hard water.
Salt and nitrite uptake. The chloride ion of salt has the desirable ability to inhibit the uptake of nitrite into fishes' blood. During shipping, fishes' excreted ammonia can form nitrite, and the addition of enough salt to make a 1% solution has been found to cut shipping losses by as much as 90% in commercial practice. Nitrite is toxic, as you know. Fish that are stricken with nitrite poisoning get lethargic. With higher levels they may gasp as if they were suffocating and die with their gillcovers open wide. The nitrite ion has the damaging habit of occupying the place on a hemoglobin molecule where oxygen ought to be carried. The resulting molecule, called "methemoglobin" carries no oxygen. Under the influence of high NO2 levels, the fish may suffer from "brown blood" syndrome or methemoglobinemia (yeah! "Me-THEME-o-Globe-anemia"). Though the main effect of nitrite is on the oxygen-carrying component of red blood cells, it's recently been shown to suppress chloride cells in the gill lamellae, which play a major role in maintaing a balance of salts. (Download the abstract of a 2002 paper read by O.T. Ferreira da Costa and M. N. Fernandes of the University of Sao Carlos, Brazil, "Chloride cell changes induced by nitrite exposure...")
How much salt should you be adding to counteract nitrite? It is the chloride ion of salt that is effective, not the sodium ion. In order to be effective, the chloride-to-nitrite ratio should be five to one. So if nitrite tests at 1 ppm, you should add enough salt (as a temporary measure) to give a chloride level of 5 ppm. This corresponds to about 8.5 ppm of NaCl (table salt), very little — a fifteenth of a teaspoon or just a pinch — in ten gallons. In fact, your water quite likely already carries this much salt, without any extra dosing at all; at any rate, your normal partial water changes will dilute out additional salt after the crisis has passed.
By the way, that useful chloride ion could perfectly well come from another source: you could use potassium chloride instead of sodium chloride, and plants would benefit from the potassium. Calcium chloride, CaCl2, has two chloride ions; though not as cheap as rock salt, calcium chloride proved in trials more than twice as effective. Check this abstract of an article reporting nitrite trials with striped bass, Morone saxatilis, (a marine fish, however). (Check correspondence at Patrick Timlin's website).
Now, try telling this to a friend who adds a pinch of salt to "ease stress." This use of salt's chloride ion does not assuage high levels of cortisol that are arguably more responsible for "shipping stress" than simple nitrite poisoning. At any rate, don't be misled by vague references to "stress." The chloride ion of rock salt simply acts as a specific block to the uptake of nitrite formed from ammonia in shipping water.
Salt and cleaning. Salt brine is a good cleaning and disinfecting agent for freshwater aquaria and accessories. To make brine, stir rock salt into warm water until some salt crystals remain at the bottom of the container and won't go into solution. "You've got brine." Articles soaked in brine should be allowed to air dry before being thoroughly rinsed.