(That does actually work!)
When I was a kid I was fascinated by the original lava lamps, but since my dad seemed to think that they were "an eruption of vomit in a bottle" I was denied the chance of getting one. Obviously this made them much more desireable, so when they became available again locally I bought one, then another, then another.......
I had made previous attempts at making a lava lamp out of wax and oil, but apart from bottles full of a wax and oil emulsion I had no success. Getting a few lamps rekindled the desire to make a lamp and I started experimenting with wax, oil and methylated spirit or surgical spirits (rubbing alcohol?). Although the concept of heating a bottle full of highly flammable substances may seem extremely dodgy, it did result in a lamp that was easy to make and worked pretty well. The alcohols were the only viable option for a liquid that would let the naturally boyant wax sink to the bottom of the bottle, and given that lamp oil (liquid paraffin / kerosene) is lighter than meths, but baby oil (mineral oil) is heavier, it was easy to adjust the density of the wax to the point of operation.
It was at this time that I really started to appreciate how important the spring at the bottom of the bottle is, since it has two very important functions. Firstly it breaks the surface tension of the descending globules of wax and allows them to merge with the main body at the bottom, and secondly it causes the wax to form a solid layer over the bottom of the bottle which allows it to heat up properly and stops it from forming a single big globule that hovers halfway up the bottle. One of the hardest thigs to achieve initially was making the spring receptive to the wax, since if it isn't primed properly the wax will simply bounce off it and form the hovering glob previously described.
In a sense the wax and alcohol formula is the perfect beginners recipe, but technically speaking it's an absolutely terrible idea from a safety perspective, and is tantamount to having a molotov cocktail on your sideboard. That said, the recipe does have some redeeming features in that it's very easy to combine the ingredients by just pouring them in and shaking the bottle until they are all mixed. If you tried that with a standard lava lamp it would result in a permanent clouding of the liquid. In the alcohol version the liquid does cloud, but soon settles into a clear solution again. As far as the priming of the spring goes, this simply involves coating it in wax so that it has an affinity to the wax in the bottle. If it isn't primed, then the alcohol will make direct contact with the metal and stop the wax from getting a hold.
The alcohol version isn't perfect. If the lamp is left on a long time the liquid will cloud slightly, but the cloudiness will form into fluffy wax when the lamp cools and will merge with the wax again when the lamp is next used. It's difficult to colour the wax too, since the alcohol will tend to scour the colour out of the wax leaving it a natural clear or white. Incidentally, you can add candle whitener to the wax to make it more opaque, but this will involve heating the wax to quite a high temperature externally to allow the opacifier to merge. Another problem with the alcohol version is that over time it may scrub the wax off the spring requiring that it be removed and re-primed. Maybe a much higher melting point wax would help make a more robust surface on the spring.
Since the alcohol version was a bit dodgy I decided to investigate the alternatives and did a huge amount of searching on the Internet to see if I could find any information. (I'm not even going to mention the raisins in lemonade "lava lamp"...DUH!) The first recipe I found was ironically the alcohol and oil version that I had already discovered for myself, but the only other recipe I could find was the benzyl alcohol one and quite frankly it was rather problematic in that it required a detergent to break the surface tension for reliable operation, and the detergent caused major clouding due to the combination of the alcohol with the water. Further investigation of the Internet found a comment regarding the original Mathmos formula of the lava wax using carbontetrachloride as a heavy solvent to make the wax heavier than water. Unfortunately that particular solvent is so environmentally unfriendly (for the exact reasons that make it great for lava lamps!) that it is not possible to buy it from my local chemical suppliers. So after a bit of further research I concluded that a common electronics industry solvent called trichloroethylene would be worth trying.
I duly bought a large bottle of trichloroethylene from my local lab suppliers (ouch!$$$) and made up a lava lamp using the solvent, oil and candle wax. This is when I discovered the importance of priming the spring and also eventually discovered that it's vitally important to have a scrupulously clean bottle and make sure that the wax NEVER contacts the glass directly without water being present to stop the wax from sticking to the glass. If the wax does take to the glass, then the lava will slide up the side of the bottle leaving a permanent waxy skid mark that acts as a preffered route for the lava for future operation (not good). I also discovered that it was essential to add a drop of detergent to the water to break the surface tension of the lava and make it go up easier, and this caused major clouding problems when the lamp was up to temperature and the oil in the wax emulsified with the water.
After a lot of messy and expensive experimentation I was pondering how to get a wax that didn't have the oil seepage problem, and I suddenly thought of how hard it was to wash Vaseline (petroleum jelly) off your hands, since soap doesn't help remove it. I duly made a batch of wax using trichloroethylene and Superdrug petroleum jelly and was rewarded by a lamp that works oustandingly well.
To make a lamp in this style requires the following... Get a good clear bottle with flat bottom and form a spring into a circle that will easily push through the neck of the bottle and open loosely across the bottom of the bottle. It's easiest just to cannibalise a dead lava lamp for these components. If the bottle has any traces of wax on the glass at all, then the only way to clean it is to use caustic soda and water. Caustic soda is violently reactive and turns grease into a soapy substance, so be very careful when using it. Heat the petroleum jelly in a non-food container and when it has melted add trichloroethylene to the point that when a drop of the mix is squirted into a jar of cold water, it sinks. Marinade the spring in the mix and then remove to cool, also allow the mix to cool and solidify. (do the heating and mixing in a well ventillated area!). Fill the bottle with water and a few drops of detergent, and push the spring through the neck so it falls into place at the bottom of the bottle, put the solidified lava mix into a plastic sandwich bag then cut off a corner from the bag and squeeze the lava into the bottle like a big gooey worm.
Heat the bottle up on either an improvised base fitted with a 40 Watt lamp or better still an existing lava lamp base, and if you're really lucky (HAH!) the lamp will heat up and start working. Most likely though the wax will stay at the bottom and even after the crucial three hour minimum warm up time, will still be rippling at the bottom. In this case just leave the lamp in a well ventillated area and the excess solvent will slowly bubble out of the lava and escape into the air until the lava is light enough to float at the top of the bottle, whereupon it can be fine tuned by adding single drops of trichloroethylene noting that you may have to let the lamp cool down to let the wax sink if the trichloroethylene shoots right through the wax and ends up at the bottom of the bottle. A small quantity of detergent breaks the surface tension of the lava and lets it ooze up easier, and the quantity used will determine the globbiness or ooziness of the lamp. Without it the wax may just sit on the bottom and bulge gently in a half hearted manner.
Patience is absolutely vital when making a lava lamp, since the lamp will only operate in a stable manner when it has reached thermal equilibrium, which means that the lamp isn't getting any hotter. A lamp that operates too early will usually stop operating when it gets up to heat, with the lava staying at the top. When the lamp has been fine tuned, then it is really important to ensure that a tightly sealing lid is put on so that no more solvent escapes, even under the higher internal pressure of a hot lamp. The lid should be removable so that final adjustments to the lava can be made, and for colouring you can add food colouring to colour the water and oil colourant (for lamp oil?) to colour the lava. Just squirt them both in the top and they will find their own way to the correct place. A plain uncoloured lamp is great since the petroleum jelly lava is naturally translucent and the lamp can be coloured using a coloured lamp allowing different colours to be chosen as desired.
The recipe is very good, but it's not absolutely perfect since the solvent may seep into the water over time initially, but in general I think the recipe is easily comparable to many commercial brands of lava lamp and definitely better than some of the cheaper brands. That said, I think that it's a lot safer and cheaper to buy commercial lamps, since the chemicals involved are extremely unpleasant and you will save a lot of hassle. (but then I know you just MUST build your own).
If you have to get rid of a batch of reject lava, then first heat the bottle to melt the wax, then pour the hot solution into a bucket or other container. Once the wax has solidified again you can pour the water away and dispose of the wax.
NEVER pour the hot solution directly down your sink drain, particularly if experimenting with candle wax based lava. If you do then the wax will form a hard honeycomb structure on contact with the cold water, and this could block your drain pipe. I did this ONCE and had to remove the sink trap to clear it out. Fortunately the cold water in the trap had solidified the wax meaning that it didn't go any further.
I actually had more luck with making a couple of bigger lava lamps since the larger quantity of liquid means that there is a much greater temperature difference between the base and the body of liquid. This basically means that the high temperature difference leads to an exagerated expansion and contraction of the lava making the effect easier to achieve at low temperatures. This has the added bonus of allowing a really simple recipe of trichloroethylene and oil to be used since the lower temperature reduces the risk of the oil emulsifying with the water and detergent.
For a big lamp you will need to use a hydrometer to measure the density of the oil and solvent, since adjustments can be much harder when the lava is a liquid. You should mix the trichloroethylene and oil until the density is just a bit higher than the water then fill your huge jar with water, drop in the spring after dipping it in the lava mix (or wax it perhaps?) then add the lava which will hopefully drop to the bottom and merge on the spring. Try the lamp as it is, but you will probably need some detergent in the water and this can be mixed with water and poured in, or alternatively if it's just dripped in it will tend to sink to the bottom and merge with the lava instead. I think having it in the water is better.
Seal the top of the lamp (as best as possible) and when the lamp is up to temperature if the lava is sticking at the top then add a few drops of trichloroethylene until the balance is reached. If it steadfastly stays at the bottom then you may have to increase the density of the water by adding a mixture of water and salt.
The larger lamps tend to be a bit more sensitive to ambient temperature, especially when they are being run on a very low wattage bulb. I've run one on a 15W lamp, a 6W heating mat and even had one that did it's stuff in the morning sunlight, although that was a bit too temperamental. One downfall of the larger lamps is that the water must be pretty sterile to reduce the risk of mould growing in the water. The trichloroethylene seems to help inhibit this though.
My final bit of advice is to restate my previous suggestion of going and buying several commercial units. They are very cheap these days and it will save a lot of grief in the long run.
Obviously there are some of you who like me MUST make something themselves to appreciate it properly. Believe me when I say that I REALLY appreciate lava lamps now.
If you follow these instructions for making a lava lamp then you do so at your own risk. Some of the materials are hazardous and should be handled with extreme care. I will not be responsible for any adverse results of lava making under any circumstances, whether it be death or big waxy stains on delicate furnishings.
Have fun and take care.