Forms - Moonmilk

This article is about deposits of moonmilk often seen in Eastern Australian caves.

Origin of Moonmilk
Tourist Do's and Don'ts
Discussion about thin wall coatings
Sharks Teeth Stalactites
Fluffy Stuff


Moonmilk is the term given to the white pasty material often seen coating walls and stalactites in caves. As a generic term, it gives no hint as to origin or composition. Moonmilk is found in many caves all over the world.
In Australian caves, I have found moonmilk in some interesting forms as well as the usual coatings as follows:

Origin of Moonmilk

Hill and Forti cite four main theories as to the origin of moonmilk:
  1. Freezing in ice caves (does not explain tropical moonmilk).
  2. A by-product of the life cycle of various microorganisms (does not explain all occurrences of moonmilk)
  3. A disintegration product of bedrock (only explains one type of moonmilk)
  4. Precipitation from groundwater in which there is an agent which prevents the crystals from growing large. This is the theory preferred by Hill and Forti.
Let me address each of the above points:
  1. Can freezing conditions cause moonmilk to deposit? This is a report of a simple experiment performed at Jenolan Caves. Tapwater at Jenolan Caves contains a high level of calcium and carbonate in solution. It is pumped from the underground creek flowing through the caves and processed with a UV light steriliser. If Jenolan Caves tapwater is chilled in a container in a kitchen freezer, the calcium carbonate precipitates first as calcite rafts on the surface of the water. They then may sink to the bottom of the container as calcite rafts. The ice crystals form at the top of the container as would be expected. The sunken calcite rafts do not look at all like moonmilk; they are hard flakes, not a soft paste.
    In a colder area, water freezing on the surface of a cave wall generally results in flakes of bedrock being dislodged due to ice crystal wedging, but this depends on the type of bedrock as to what type of disintegration occurs (see below). In mountainous areas, rock flakes due to frost wedging are common on the surface, and the flakes are not moonmilk. The grinding of rocks by glaciers produces rock flour, but this is made of fine pulverised rock and does not resemble moonmilk. Yet moonmilk does occur in icy caves. Perhaps it has nothing to do with the ice.
  2. Is moonmilk due to the product of the life cycle of various microorganisms? Some is (see Phillips and Self). What seems to be happening is that calcite is deposited as a byproduct of microorganisms either as a coating over the microorganism (eg travertines), as a deposit inside the organism such as in tubes of fungal threads, or as a fairly conventional (mineral) deposit of crystals very close to the microorganisms.
  3. Is moonmilk due to a disintegration product of bedrock? This of course depends on the type of bedrock, and what is the substance causing it to disintegrate.
    One common cause of bedrock disintegration is bat guano. The disintegrating substances in this case would be various nitrogen-based and phosphorous-based acids. These act not only on the bedrock but also on other speleothems and cave deposits, forming the multitude of bat-guano derived minerals. Some of these may have a moonmilk appearance. Also some of the acids may be quite weak, causing only part of the bedrock to be corroded and leaving larger grains and bioclasts in the bedrock untouched.
    Other substances and causes of bedrock disintegration in caves may include:
    • The weathering of exposed pyrite in cave fill and bedrock to form hydrated iron oxides and sulphuric acid. This usually results in an absence of bedrock rather than a deposit of moonmilk.
    • The precipitation of gypsum causing bedrock to wedge apart (crystal wedging); salt wedging by the precipitation of various other salts such as halite, nitre, etc in dry caves. This usually results in small bedrock flakes and dusts rather than moonmilk. One cave in which I was surveying (part of the Wollondilly tourist cave at Wombeyan, NSW) has an interesting fine calcite sand. This appears to be wedged from the low roof. The bedrock in this case is marble and the area was near a bat roost.
    • Fault zone rock flour is quite distinct and again does not have the cottage cheese texture that the usual moonmilk has. It should be classified as "rock flour" in much the same way as glacial debris.
  4. Is moonmilk precipitated from groundwater in which the crystal size is kept small?
    • What if precipitation is rapid? Rapid precipitation of substances will generally result in small crystal sizes. This occurs near cave entrances where precipitation occurs due to both outgassing CO2 and evaporation of water. It also occurs where gypsum is being produced (common ion effect).
      It does not explain the pasty types of moonmilk, and does not explain why stalagmites and stalactites formed in cave entrances are hard and crystalline and quite unlike moonmilk. Therefore I don't think rapid precipitation in itself is what causes moonmilk.
    • What if the calcite grains are prevented from touching one another? This could occur if the grains were formed inside an organism (such as fungal threads as is proposed by Phillips and Self). This could also occur in an organic fabric or colloid, such as in microbial veils. It could also occur if the calcite was deposited as a by-product of micro-organisms' activity as each calcite grain would be associated with a specific set of microbes.
      It should be kept in mind, however, that many microphotographs of moonmilk show that it is composed either of fine rods with a diagonal texture, or fine particles of calcite (sometimes hollow, skeletal crystals). This is difficult to produce from purely inorganic sources.
      Calcite grains with a diagonal texture are common (eg in fossil bioclasts) however the texture is simply common pressure twinning. It is the rod-like aggregate that I am arguing has a biological source.
      Further examples are shown in discussions on lublinite.
    • One way to decide whether the moonmilk is deposited as a result of organic activity would be to take some samples, verify that they are still depositing calcite in controlled conditions, then adding small quantities of various antimicrobial agents such as bleach, antiseptic solutions, etc and see if the activity stops.

Tourist Do's and Don'ts


Further notes about thin wall coatings

The thin deposits of white moonmilk associated with batty areas may be simply the results of chemical action from guano on bedrock.
The zone of "moonmilk" appears to be only about 0,5 to 1 mm thickness and its substrate is limestone bedrock. It does not occur in well-washed areas (as it is easily washed off - as I found when preparing samples). It also occurs in areas in which cave fill (mud and gravel) has been deposited, and again in these places it occurs more on the limestone bedrock and less on the fill. One of my samples of "moonmilk" deposited on clay comprises major quartz, minor lepidolite, albite, kaolinite-1Md and aragonite. In some areas it is associated with black deposits. A sample of this white powder from a cave in the northern area of Jenolan was analysed by Ross Pogson at the Australian Museum using XRD and found to be mainly calcite with minor quartz, very minor magnesite and minor hydroxylapatite. The site is associated with dolomitised limestone, fault zone breccia, bat guano, clay and aragonite speleothems.
Possibly what is happening is the lime mud in the bedrock is being corroded first, leaving small bioclasts such as peloids and echinoderm fragments relatively intact.

Further notes about Sharks Teeth stalactites

This is a small stalactitic form that occurs near cave entrances and resembles sharks' teeth.

I have seen this form in a cave at Bats Ridge (Victoria) where the bedrock is aeolian calcarenite. It appears to be made of a mixture of moonmilk minerals and something with chlorophyll, eg cyanobacteria. The outer side of these little stalactites is greenish whereas the inner side is white. The stalactites have a pronounced flatness as though to maximize the amount of light falling on them. Each "tooth" is about 1 cm wide, 2 cm long and about 5 mm thick at its base. The "sharks' teeth" are deposited en echelon so as to form a stalactitic aggregate. The cave also has large tray-like stalactites of moonmilk, and dryish moonmilk covers most walls.
The sharks' teeth form clumps which look like normal stalactites from a distance.
sketch of sharks teeth stalactites

Possibly what is happening is this:
Aeolian calcarenite is a very porous rock made of grains of calcium carbonate and silica blown into sand dunes and cemented. The ones at Bats Ridge are Pleistocene in age (White, 1995). The porosity of the rock allows seepage water to moisten a larger area of the cave roof than would occur in harder, more consolidated limestone. At Bats Ridge, most roof surfaces are covered in moonmilk to some extent. White states that the moonmilk is needle form calcite, with a high moisture content between the fibres.
In a large chamber in the cave, large (about 2 m long and 0,5 m wide) stalactites of moonmilk occur, presumably at those parts of the roof with more seepage water than elsewhere. The sharks teeth stalactites only occur near one of the cave's entrances on the sunny side.
One possibility considered was condensation of moist warm air at the entrance during the cooler months, however this should lead to forms seen in other caves and not specifically sharks teeth stalactites.
Maybe during the cooler months (winter is the wet season) there is more moisture available to feed the moonmilk stalactites. During summer, the sun passes higher overhead but in winter, the area where the sharks teeth stalactites occur may be directly illuminated with the lower angled sunlight. This would allow calcite to deposit more rapidly where the sunlight falls on wet areas of the sharks teeth. This is similar to the discussion regarding crayback stalagmites and spelothems that appear to lean towards the light.
The caves are relatively shallow and the roof of the cave is well within the zone in which soil bacteria thrive. The vegetation above is sclerophyll, with Banksia and low Eucalypt. Moisture content must be fairly high as leeches are a hazardous "feature" of the area.

Futher notes about Fluffy moonmilk

There are at least three occurrences of a rather fluffy variety of moonmilk which I have encountered in NSW caves.
Thin fluffy coatings
In the Wollondilly (tourist) cave at Wombeyan Caves, there is a deposit of fluffy moonmilk covering one wall in an area called "The Loft" where there is at times a fair bit of air moving from one part of the cave to another. The deposit is about 5 mm depth and covers an area of about 3m high by 10m wide. It is developed on a mixture of sand and igneous rocks which have been washed into the (marble) cave. There is some evidence for an old bat guano deposit in the area, based on the corroded appearance of stalactites and stalagmites nearby. The deposit is mostly white, with patches of cream and light orange. In situ, the deposit glows bluish under long wave UV light. sketch of fluffy moonmilk on wall
Under the microscope, fresh fluffy material looks like a thatch with bits of what appear to be quartz grains. After dessication, the material becomes a fine white powder and loses its thatch properties. Before I was able to test a sample of the material, I guessed that it might be a mixture of needle-form calcite, silicates and maybe phosphates, maybe a type of soil developed as a result of bacterial action on bat guano, volcanic rock, lime-rich water and the drying effect of air movement. The guess was close. Results of tests show the sample comprises needle-form calcite, ardealite, aragonite, vaterite, quartz, several phosphates and silicates. For a detailed analysis of the material, see Needle Form Calcite in Wollondilly Cave.

A similar kind of fluffy deposit occurs on the surface of hemispherical deposits at Jenolan (NSW), locally called "potatoes" after the fancied resemblance to the vegetable. I first noticed this during the survey of a cave in the northern end of Jenolan.
A very small quantity of the fluff was rubbed between the fingers. Like other forms of moonmilk, it felt pasty like cottage cheese and nothing like fungus.
I thought that it may have been made of the same material as the "potatoes", only the "potatoes" are older and more consolidated. This will be discussed separately in another article.
sketch of fluffy moonmilk on hemispherical deposits

Possibly the fluffy deposit is the active growing part of the hemispheres. Where this deposit is found, the cave is dry and the air is still although the humidity is high. I have seen examples of this form in several caves at Jenolan, including the showcaves (Lucas and Chifley), and in one cave at Wombeyan. In each case, it occurs with a suspected ancient bat guano deposit.

"Fluffy Stuffite"
The unofficial name for the substance was coined by ANU student, Mr I. McCulloch during a mineralogical survey of the Chifley cave with members of the Australian Museum in 2000. This most intriguing of the fluffy deposits at Jenolan Caves occurs in the Wilkinson Branch of the Chifley (show) Cave at Jenolan.
In 1899, John Mingaye (chemical analyst for the NSW Goverment Department of Mines and Agriculture) analysed a sample of this material. He wrote:
Two samples of this substance were received. The first, which weighed 1½ grammes, Guide Wyburd states, "was compressed into a small match-box, and would fill your hat in its natural state."
sketch of fluffy stuff on rocks

"It is so light that, when you blow at it, it falls off the roof and sides like snow." It is stated to occur in one cave only - that is, at the end of the Wilkinson Cave. The fungoid growth appears to be very rapid, for since the sample was received it formed in large quantities, and has the appearance of very fine wadding.

The analysis showed the bulk of the material to be calcium carbonate (72-76%), silica (13-17%), magnesium carbonate (5-6%) and small quantities (around 1%) of moisture, alumina, ferric oxide, organic matter. There were traces of strontium carbonate and phosphate.
There does not appear to be hatfuls of the material presently in the Chifley Cave, however the material that I have been shown is certainly not dense. Maybe its development is seasonal.
Some of it has been analysed recently by the Australian Museum and was found to be mostly calcite and a small quantity of silica as quartz. The analysis did not indicate what form the calcite took, i.e. whether it was needle form calcite, fine crystalline calcite or other form. The material that I saw took the form as sketched in the figure, with some as a loose ball on the soil, other blobs in nooks between rocks and other pieces on top of rocks. This part of the cave is relatively dry, although vermiculations under some of the roof boulders would indicate that at times it can get damp. The air is still and humid.
Possibly the material is formed from a similar substrate to the "potatoes", i.e. a very old guano pile mixed with silt and kept moist with lime-rich seepage water. I suspect there is a little ecosystem of soil microbes involved, keeping the calcite together as a thatch of light fluffy form, with kaolin nearby in the cave acting as a moisture stabiliser. It would be interesting to investigate this further.

Here are my photographs of the fluffy material in the Wilkinson Branch of Chifley Cave.


Some brief reviews are included with the references.

Cosmetic update January 2006. Originally written January 2002, updated 17th February 2006.

home speleothems cave zone by form cave minerals pseudokarst conservation articles about jill

By Form

Moonmilk Sharks Teeth Fluffy Stuff "Potatoes"