December 27, 2003, a crisp and sunny day, perfect for a hike.  Greg and I went to Big Ridge State Park in east Tennessee.  Soon we came upon a cut off log with a giant icicle extending out of the lower end. It must have been below freezing for many days to get an icicle of this width and length.  But, it must have been warm enough in the log and the soil to keep the water flowing so the icicle would keep getting a supply of water. 

A little later I noticed white puffs off to my right, as though someone had dropped tissues.  I looked closer and saw that these objects were ribbons of ice.  They were not growing out of the ground as wedges of ice do.  These ribbons seemed to be hanging from branches.  I had never seen anything like this before.  I took a few photos and went on my way.  Since then I have studied these photos trying to figure out what I saw. 

It was almost noon when I found these and the temperature was now above freezing.  The more I review this photo I am certain the ice has slid down the twigs from the original position.  The little patch of ice on the lower twig on the left makes me think the larger ribbon of ice was once located here. 

In the example below you can tell the size of the ribbons by comparing them to the Oak leaves on the ground.  I first thought this ribbon of ice radiated out from the node at the junction of the stems.  I now realize it grew, from the stem and rotated around the junction of the stems.  This same process seems to have occurred in the second ribbon at the bottom.  I am not ready to explain what produced the tight spiral in between these two fan-shaped ribbons.

Below is one that seems to be curled up along one edge.  That edge does not appear to have broken off, but perhaps it was broken and differential melting has produced this scalloping.

We broke off a piece and looked through it.  These ribbons of ice were very thin, as shown by the piece in Greg's hand.  In this example the parallel bands look like growth rings, but it likely these striations are formed by variations in the width of the slit from which water emerged and froze.

Below is another example, in this case above a non-leafy surface.  This demonstrates leaves are not a required component of this process, but the stems are present.

As I remember all of these ice ribbons were found along a north-facing slope which would be cooler and would delay melting.  In December 2004 I returned to the Park to see if I could find the place where I saw these ice ribbons the year before.  I did not find that place but further along the trail I found some ice ribbons or frost flowers in somewhat different forms.  I have two pages showing what I found in 2004.  Below is one such photo.

In 2005 I found such ice formations in northern Kentucky.  In fact I found them in many places that November morning.  Below is one image from my trip through northern Kentucky.

The weekend after Thanksgiving 2005 I was in central Virginia.  It was a cold and frosty morning and so I went exploring.  By golly I found outcrops of needle ice and then I found a few ribbons of ice, as below. 

Check out the collection from central Virginia. for it is different from I had found elsewhere. 

In fall 2004 I decided to put my photos on the web so a colleague could see them.  But first, I turned to the Web to see if I could find any information about these strange ice formations.  I searched on a number of terms and found a few web pages showing ice formations similar to what I had found.  More recently I have compiled a page of links to other sites showing ice flowers.

That fall I posted my web page for the world to see.  A number of persons saw my pages and sent me email.  I have corresponded with a number of these persons and I have learned in the process.  I call these ice ribbons and ice flowers based on their appearance.  But the term frost flowers is commonly used while there is the occasional use of ice fringes, ice filaments, and rabbit ice.  

I also employed bibliographic search tools to find out about frost flowers.  I found a few popular articles describing these unusual items, but could find no authoritative papers explaining these ice formations.  I found two plant species associated with the name frost flower:  Dittany, or Wild Oregano, Cunila origanoides, and White Crownbeard, or Frostweed, Verbesina virginica.  Thus, I concluded this was a concern for botany.  I showed what I had found to many colleagues in the biological sciences and only one had personally seen these frost flowers and he thought they occurred on only Cunila origanoides.  He said he was sorry he never took this on as a research topic. 

Ice Flowers and Needle Ice

One person in Alabama sent photos, three showing ice flowers and one showing needle ice. He suggested they were related, but I rejected that association because I was convinced it was a botanical issue.  Then in Virginia in 2005 I found ice flowers and needle ice in the same immediate area, as shown above.  By golly, he was right. 

I went back to other correspondence and found another statement relating ice flowers and needle ice.  I then turned to bibliographic search tools and found there is a well developed professional literature on needle ice.  In 1988 D. W. Lawler published “A Bibliography of Needle Ice” in Cold Regions Science and Technology (15: 295-310).  There are 267 items in his compilation going back to 1824.  I delved into many of the papers cited there. 

Stories from the 19th Century

Of particular note was the letter of J. F. W. Herschel dated January 12, 1833, in Philosophical Magazine, 3rd series, 110-111, entitled: "Notice of a remarkable Deposition of Ice around the decaying Stems of Vegetables during Frost."  He wrote that years before he had found ice ". . . to incrust the stalks in a singular manner in voluminous friable masses, which looked as if they had been squeezed, while soft, through cracks in the stems."  Then on January 11 he found a similar formation of ice which he described as ". . . seemed to emanate in a kind of riband- or frill-shaped wavy excrescence, -- as if protruded in a soft state from the interior of the stem, from longitudinal fissures in its sides, . . . the structure of the ribands was fibrous, like that of the fibrous variety of gypsum, presenting a glossy silky surface"  He goes on to make additional observations about the ice and the atmospheric conditions when these formed, all consistent with what I have observed more than a century and a half later.  He ends with "What share the physiological functions of the plant may have in the phaenomenon, or whether it be connected with the vitality of the stem at all, it is for botanists to decide." 

Herschel’s paper prompted Professor Rigaud of Oxford to recall his observation in 1821 of similar ice formations on a recently built stone wall.  “The portions of the ice (with a single exception) were formed at the edges of the stones,-- indifferently at the tops, to bottoms, or the sides, but the curvature was uniformly turned inward from the mortar itself, in which case the threads of ice were formed in an horizontal line, and I think (for of this I made no memorandum) parallel to the layer of the mortar.”  Philosophical Magazine, Feb. 2, 1833, 190-191.

In the same journal (The London, Edinburgh and Dublin Philosophical Magazine and Journal of Science, Third Series) of May 1850, 329-342, John LeConte, M.D., of the University of Georgia wrote about many instances of observing frost flowers and needle ice in Georgia.  He quotes liberally from Hershel (1833), for he appreciated the words as I do.  LeConte also produced some lovely descriptions such as ". . . the traveler who passes along the level roads of this region soon after sunrise cannot fail to be struck with the remarkable accumulations of voluminous friable masses of semi-pellucid ice around the footstalks of the Pluchea which grow along the road-side ditches.  At a distance they present an appearance resembling locks of cotton-wool, varying from four to five inches in diameter, placed around the roots of plants; and when numerous the effect is striking and beautiful." p. 330.  He observes that in some cases ice had formed on the same plant on consecutive nights "when the wood was not rifted." p. 332.  He also makes observations about what we now call needle ice and argues "that both of the phaenomena must be referred to the same cause.  If we admit an identity of cause in the two cases, it is obvious that it must be purely physical . . ." p. 336.

In 1880, the Duke of Argyll writing in the January 22 issue of Nature described such ice formations and asked for a scientific explanation of this phenomenon.  In the January 29 issue three persons weighed in with opinions based on what they had seen.  In the February 19 issue the Duke responded.  In the February 26 issue one of the earlier writers and a new writer offer their suggestions based on ice formations they have seen.  All of these contributors were from what we now call Great Britain.  In the April 22 issue Wm. LeRoy Broun from Vanderbilt University describes the growth of needle ice and compares it to the ice formations observed in Europe.  Broun does not reference LeConte who was also from the southeastern U.S.

A report on the meeting of the Physical Society in Berlin, in the March 13, 1884, issue of Nature includes a discussion from Prof. Schwalbe on flowers growing from rotten twigs lying on the ground as ". . . ice-excrescences of soft, brilliant, asbestine appearance, and uncommonly delicate to the touch. . . ."  Prof. Schwalbe brought some of these withered and rotten twigs with him to Berlin, and it was in his power to produce on them at any time the phenomenon just described.  For this purpose all that was needed was thoroughly to moisten the twig, in such a manner, however, that no water dropped off, and then to let it cool slowly in a cold preparation.  Ice-excrescences also appeared of themselves on twigs lying in the garden whenever the temperature fell below 0 degree C. in the night."  (p. 472)  He made reference to the explanations of LeConte, 1850.  I originally assumed these were ice flowers but subsequently I find others reporting this phenomena on rotten twigs.  I have a separate web page on these ice formations, which are called Haareis, Kammeis, Pipkrakes and Silk Frost. 

In a later editions of Nature there were a series of letters reporting on ice formations and reacting to earlier letters.  Most of these reports relate to what was obviously needle ice but in the January 1, 1885 issue B. Woodd Smith tells of a friend who “. . . picked up a  piece of a dead beech-branch which was covered with filamentous ice, such as is described by the Duke of Argyll and others.”  This person found the ice reappeared again the next morning when it was left out over night.  (p. 194). 

I marvel at the observations and vocabulary of these authors, writing more than 120 years ago. Do we know more about frost flowers and needle ice than they did back then?  We certainly know more about needle ice, as evidenced by the many articles cited in Lawler's bibliography.  Among these papers are reports of scientists growing needle ice and controlling the rate of growth. 

Lester F. Ward, “Frost freaks of the dittany,” The Botanical Gazette, 1893, observed these near Accotink, VA.  He has stylized drawings of the stems and ice ribbons.  He writes that the article belongs in a botanical journal because of all of the plants in the area the ice formations occurred only on Cunila Mariana, dittany.  He and a colleague tasted the ice and inferred from this that the water was not “. . . distinguishable from pure distilled water . . .” p. 185  Ward said he was able to find no records of others observing this phenomenon.  He also wrote: “It is possible that this is the first time that Cunila Mariana has been discovered to be a frost-weed.  At the time the discovery was made it had quite escaped my memory that Helianthemum Canadense behaves in a similar way.” P. 185. 

Ward references Gray’s Manual, 1848, as describing such ice formations on Helianthemum Canadense, or frostweed.  As similar statement appears in the Eighth Edition of Gray’s Manual of Botany, 1970, on p. 1017 and notes that Helianthemum Bicknellii is also called frostweed. 

Ward also refers to the book Sharp Eyes by Wm. Hamilton Gibson, dated 1892.  I got to see the 1904 edition of this delightfully illustrated work, where Gibson writes about what can be seen in nature every week during the year.  The November 3d entry is The Frost-Flower as it appears on Helianthemum Canadense.  He notes it has three distinct types of blossoms during the year.  In November “the flower from which the plant is named, but which few people ever see.  Almost any morning during the past week, after a severe frost, would have shown it to us among the stubble where the plants area know to grow, glistening like specks of white quartz down among the blown herbage close to the base of the stem.  It is a flower of ice crystal of purest white which shoots from the stem, bursting the bark asunder, and fashioned into all sorts of whimsical feathery curls and flanges and ridges.  It is often quite small, but sometimes attains three inches in height and an inch or more in width.  It is said to be a crystallization of the sap of the plant, but the size of the crystal is often out of all proportion to the possible amount of sap within the stem, and suggests the possibility that the stem may draw extra moisture from the soil for this special occasion.  The frost-flower is well named.”  The sketch accompanying this text shows an blossom of ice in one image overlaid on top of the plant in full bloom in summer. 

And in the Early 20th Century

Prof. Cleveland Abbe, “Ice Columns in Gravelly Soil,” Monthly Weather Review, 1905, 157-8, writes about needle ice and references LeConte, 1850.  He notes that “Only once have I seen the corresponding phenomenon of a thin ice sheet of parallel ice columns exuding from a vertical crevice in the bark of a tree, many beautiful examples of which are given by Professor LeConte and Sir John Herschel.”  Abbe rejects the explanation of LeConte and offers his own suppositions.  He calls for someone to repeat the process in the physical laboratory.  There is a concluding remark “This explanation of the growth of hollow columns of ice in gravelly soil applies with slight changes to the hollow stems and plates of snow crystals.  The whole subject of the growth of crystalline forms needs elucidation.” p. 158. 

Then I found Coblentz, “The Exudation of Ice from Stems of Plants” in the Journal of the Franklin Institute, 589-621, Nov 1914.  Coblentz was a physicist working for the National Bureau of Standards in Washington, DC.  In 1913 he found some frost flowers in Rock Creek Park and started observing what he saw.  He systematically made observations, took notes and carried out many experiments with the help of colleagues.  He found that in the mix of plants in the rocky slopes, the ice flowers occurred only on Cunila mariana, or Dittany.  He cut off stems and inserted them in moist soil, test tubes and crucibles.  He reported on how rapidly water moved up the dry stems of Dittany and was able to grow ice ribbons, what he also called ice fringes and ice filaments.  He showed that the roots of the plant are not necessary for the formation of ice, nor is the outer bark. He applied different treatments to the stems and showed that the water for the ice comes from within the stem and is not deposited from the air. He noted one ice flower weighed 5 grams and observed that many were of this size. 

Other than Schwalbe (1884) in the Harz Mountains, Coblentz appears to be the only person who has systematically grown ice ribbons and reported on them.  (Mention should be made of Bruce Means, The American Gardener, Jan/Feb 2005, 36, who mentioned he grew beautiful ice flowers on stems of flat-seed sunflowers.) Coblentz has many photos of the ribbons he grew, but the copy of his paper I saw was on microfilm and the photos were not very clear.  He has sketches of the setup of his experiments and of some of the ice ribbons he grew and thus Coblentz demonstrates the ice ribbons are a product of a physical process.  This is an important paper to understand the nature of these ice formations because today many web sites attribute the ice to frozen sap.

The Environments in Which Such Ice Formations Are Found

But the process works only on the stems of particular plants.  Coblentz carried out his observations on Dittany, Cunila mariana, which is now known as Cunila origanoides.  Many persons have observed that Frostweed (Verbesina virginica) or White Crownbeard, produces such ice ribbons.  And, others write about Frostweed (Helianthemum canadense) and note it produces these ice formations.  I have seen many examples of ice formations on Dittany and White Crownbeard, but have not been able to find any photographic images of such ice on frostweed Helianthemum canadense.  But, I will continue to look.

LeConte (1850) wrote “The plants on which I have observed it are two species of the genus Pluchea of DeCandolle, or Conyza of the older botanists, viz. Pluchea bifrons and P. camphorate.”  I have not been able to relate these species names to ice formations nor have I been able to find new species names that were formerly known by these names.  I assume LeConte observed his ice formations on one of the three species we know about today.

Herschel (1833) stated that he first saw such ice “. . . round the roots and stumps of some dry and decaying thistles.”  The second time he found such ice on stumps of a bed of heliotropes but these stumps included stems as he describes in his paper.  I cannot tell what specific plants were involved in the formation of these ice formations. 

A. Hillefors wrote about “Needle Ice on Dead and Rotten Branches” in Weather, 1976 (31, pp. 163-168).  He observed these in Sweden.  The photos I saw were on a poor quality photocopy.  Then, in early 2006 I received an email and three photos from Wales showing flowers of ice growing from rotten branches.  In these photos the ice appears to be a little more needle-like but it is quite attractive.  Subsequently, I have received photos of such ice on pieces of rotten wood on the ground from a number of places in Europe and in one place in the USA.  These ice formations are not the same as Ice Flowers on stems extending up from the ground.  But, they occur under similar weather conditions.  I have a page showing these ice formations   These have been called Haareis, Kammeis, Pipkrakes and Silk Frost.  I am certain these are the same ice formations Schwalbe wrote about in 1884.

Interestingly, while the ice formations based on dead and rotting wood are mostly from Europe and the ice flowers based on the stems of plants are mostly in the USA, there is a photo of a classic ice ribbon on a stem on the cover of the Journal of Glaciology (1993, 39:132).  This photo was taken in northern India at 16,000 feet.  So, I must be careful about making generalizations about what occurs where in our big world.

In December 2006 I received an email from The Weather Doctor Kieth Heidorn who referenced his web page where he has two photos of long and thin ice ribbons that formed on a metal fence rail.  The photos came from Sheryl Terris of Vancouver Island, Canada.  She is quoted as saying she gets these ice formations year after year on the same gate. These appear to be consistent with the observations of 1821 of similar ice formations on a recently built stone wall.  With her permission I put together a web page showing her great photos.  Then I experimented and was able to extrude ice that is very similar to Sheryl's photos.

In summary, we know the formation of frost flowers, ice flowers or ice ribbons is a physical process, not related to the growth of a biological organism.  These ice formations occur on the stems of a few species of plants and on certain pieces of rotten wood on the ground.  I thought I could count the numbers of species on which these form, and then I hear from a woman in central Tennessee who found these on New York Ironweed Veronia noveboracensis.  She convinced me she knew her plants and thus I added the stems of another plant on which these occur.  And, through these web pages I have heard from others.  As of March 2008 I believe there are at least a dozen plants on which these occur.  I thought they occurred only on dead stems, but last fall I saw them form on stems that we alive.  

There are probably more.

These form when the water in the soil remains above freezing while the air temperature falls below freezing.  The conditions would be optimal on clear nights with no wind, when the dominant cooling process is net radiation.  This fits the description of many reports of observations of ice ribbons, but, there are reports that these also occur under quite windy conditions.  See Tim Ernst:  http://www.cloudland.net/Nov03journal/Nov03Journal.html  for the dates of 11/24/03 and 11/25/03.  After observing these many times during the fall and winter of 2007 I know the only criteria is air temperatures below freezing and a soil temperature above freezing. 

On the Processes of the Formation of Ice Flowers, Frost Flowers and Ice Ribbons

What is the process that leads to the formation of these ribbons of ice?  The stems of these plants and certain rotten pieces of wood serve as conduits to transmit water from the soil which is above freezing to a surface which is below freezing.  (I cannot conceive how a fence rail could stay above freezing while the air outside is well below freezing.  I'm thinking about that.)

Charles Knight of UCAR offered an explanation in an email (2 Feb 05): “The reason this sort of thing happens is that ice cannot grow in supercooled water through very small openings, but water, of course, can flow through the openings.  So one can have an ice crystal in contact with a wet porous material, and it grows at the contact, pushing the ice crystal away while the supercooled water flows to the contact.”  Based on this there is another necessary condition for the formation of the ice -- the dew point temperature of the air must be below freezing and the temperature must fall to the dew point, or more properly the frost point.  When the temperature falls to the frost point, ice will start to form on surfaces.  The crystallization of the ice should be sufficient to start the growth of ice flowers or Haareis, etc.

Knight pointed me to the paper by Ozawa and Kinosite, “Segregated Ice Growth on a Microporous Filter” Journal of Colloid and Interface Science, (132: 1, 1989).  They define ice segregation as “. . . the phenomenon in which, through freezing, ice grows out from moisture-containing porous material such as soil.” (p. 113).  They conducted experiments growing ice over many hours while making very detailed measurements.  They found the water to be supercooled and colder than the surface where the ice forms. But, in their experiments they started the freezing process by introducing a droplet of ice on the surface of the porous filter.  They show that the latent heat of fusion flows from the surface of freezing to the supercooled water.  They conclude their paper by considering these processes in soils and the formation of needle ice and/or ice wedges.  They do not seem to be aware of ice ribbons or frost flowers.

So, what we observe when we see ice flowers or frost flowers is ice segregation along the stem of a plant.  When the supercooled water penetrating through the stem encounters the first crystals of frost the supercooled water turns to ice. That process continues and new ice is added at the stem-ice boundary.  This pushes the old ice out and away from the stem.  If the stem is not ruptured over the night, then the process can go on continuously until the ambient air temperature rises above freezing or there is a limit to the supply of water in the soil.

I have observed that such ice flowers will reoccur on subsequent nights showing that the water in the stems does not always freeze.  And, it is known that such ice formations may occur in the same area over many nights in late fall and winter. 

In fall 2007 I observed the rates at which such ice forms on stems of White crownbeard in my own yard.  I show this by a sequence of three photos over one night.  Forrest Mims III has a better record with a time-lapse video on his Youtube site. 

Thanks to the wide reach of the Internet, we are learning more day by day.  If you have observed the growth of these ice flowers, ribbons or whatever you call them and can add insight into the process, please let me know.   jrcarter@ilstu.edu

one of the pages of Dr. James R. Carter