After the excitement, for me, and the interest from readers in the ice spike featured in my previous post, I thought I’d write a little more about these unusual natural phenomena.

Ice spikes, or my earlier ice vase, below, from 2013, can only form because of the very unusual property of water that, as a liquid, it expands as it freezes. Few other materials in nature do this. Bismuth and Antimony, two related metals, also have this unusual property, but we’re unlikely to be able to study spikes with these, since their ‘freezing’ points are 272 and 631 degrees C respectively! 

Two American physicists, Liu and Libbrecht, (L&L) working at the California Institute of Technology in 2003, studied the creation of smaller artificial ice spikes in freezer ice cube trays of water, and concluded that the probability of spikes ever developing is most likely around minus 7 degrees C, and is also more likely if using impurity free distilled water, and ideally with some air movement above the tray in order to speed cooling.  Click here for their study. By filming inside their freezer, they established that the quite tiny ice cube spikes grew to their maximum heights surprisingly quickly, in between 3 to 10 minutes.

Close neighbours to us, had measured the minimum temperature on the night when our spike formed as being about the same ‘magic’ minus 7 degrees C. Under such suitable ‘supercooling’ conditions, where the temperature falls very rapidly at dusk, ice crystals which are classically hexagonal in profile, will begin form in a particular way, both spreading rapidly across the water surface and also around the base of a suitable water filled receptacle – the tarpaulin with us. The complete ice structure removed from the tarpaulin, below, gives an idea of how shallow the water pool was, and the length of the spike – estimated at about 190 mm.In addition, the ice crystals will grow down into the water, hanging like curtains, and also grow up from the base like walls. Depending on the angle these crystals start out at, they will probably not form vertically within the main body of water.

It’s been noticed that natural ice spikes like mine, and ice vases, usually have a triangular base, as in the case of my ice spike photographed from the tip down towards its base (above). This is probably because the hexagonal ice crystallites will eventually all tend to join together. If this process were as well controlled as with the efficiently constructed hexagonal wax cells in a honeybee hive, there would be no gaps between crystals. However, since ice crystal formation is a more random affair, then the potential for occasional triangular gaps to be created in the surface crystal sheet exists. At some point, under optimum conditions, the pressure of the encroaching ice crystals on the still cooling, but critically, expanding, unfrozen water below the surface, might force some of it out through just such a triangular gap.

The water will then start to freeze on contact with the colder air, but if the water is being forced out fast enough, this process can fairly quickly lead to a growing spike, or vase with peripheral walls, and a central water filled channel. Constantly extending at its tip, until either the water in the central channel freezes, blocking further water flow up it, or the supply of still liquid water in the centre of the ever thickening, ice lined ‘bowl’ is exhausted. L &L’s paper has a simple diagram illustrating this process, known as the Bally-Dorsey model, named after the two, independent scientists, who first proposed this explanation in the early 1900’s. L&L also suggest that water impurities impair ice spike formation, because they aren’t incorporated into the ice crystal lattice formation, so will tend to concentrate in the water droplet at the tip of the growing spike. In so doing, they will lower this bit of water’s freezing point, causing the spike to grow so slowly that water will freeze over completely lower down in the tube structure, before it spills out from the tip, and halt the spike growth process.

So, there you have it!

At last, I think I can really grasp how these amazing and rare structures form, and I hope that I’ve been able to explain it to any readers, in a more comprehensible way than last time. This now presents the new challenge of setting up several suitable, rain water filled containers, (and presumably our Welsh rain is wonderfully pure) in advance of any future predicted temperature plunges, and then making regular visual checks of the freezing surfaces, in order to allow the possibility of filming the growth of a natural spike, should it begin to form. Liu and Libbrecht reckoned about 50 micrometres of ice spike growth per second, so the whole process of spike formation is probably over surprisingly quickly. By my reckoning, this would equate to only about an hour for a spike, like this one, of 190 mm. How extraordinary!

I’m certainly up for another ice challenge. But we’d need some more proper winter weather to ever confirm this.

And need to wrap up well.

So for any other readers who live in a suitable climate, there’s a challenge too. And to any academics, how about creating a mathematical formula, or computer model to explain with more accuracy, the variables which are critical for allowing such amazing structures to appear?

Meanwhile, I’m now officially launching my Welsh Historic Snowdrop Hunt, (WHSH), after a trip to the National Botanic Garden of Wales (NBGW) this last weekend for another look at their ‘drops, and more particularly to hear Naomi Slade give a talk about designing with snowdrops in gardens.

Naomi has quite recently written a book, inspired by a long term love of these tiny white flowers – “The Plant Lover’s Guide to Snowdrops”. Beautifully illustrated, and full of interesting information, its other big selling point is that it describes the featured snowdrop cultivars according to both flowering season, and garden worthiness, which are two of the most critical parameters for anyone considering adding to, or even beginning, a snowdrop collection in their garden.

A couple of points raised during the talk, and Q and A afterwards, convinced me that I should ‘go public’ with my fledgling Welsh Historic Snowdrop Hunt (WHSH). Several observations I’ve made over recent years are pertinent to this project:

Firstly, many local Welsh ‘naturalised’ snowdrops are often early, and vigorous. This made an immediate impression on me back in the early 1990’s when we started to travel down from Bristol in late winter – the Welsh snowdrops were always bigger and more precocious than their Gloucestershire cousins.

Was this just because of local conditions, or were they genuine ‘good doers‘? It turned out that when our first snowdrops were moved from Bristol to Wales, they still lagged the indigenous ones in time of flowering and size.

Many local colonies are also clearly very prolific seeders, spreading in areas with no helping hand from any gardener. Indeed often thriving on neglect.

Secondly, in spite of there being nearly 2,500 named cultivars, none have, as far as I know at this time, genuine Welsh provenance. (E.O.A.) Why is this?

There is G. ‘Welshway’, but this was named after a house near Cirencester. Then there is the lovely G. ‘Mrs. McNamara’, which we grow, and which is named after Dylan Thomas’ mother-in-law. But taken from her garden in Hampshire, and possibly directly imported from overseas to grow there. Many snowdrops are from much drier parts of the UK, and some even from overseas. Most of the named cultivars will be just seedlings – natural or artificially created hybrids of species growing close to each other, or even single species variants. Perhaps some are even spontaneous mutations. Many are excellent garden plants, like G. ‘Atkinsii’, above but probably just as many are much more fussy, and struggle to survive or flourish here in our garden in West Wales. This is no doubt in part due to our much higher rainfall – a lot of named cultivars originate in East Anglia, for example, which receives about a quarter of our annual precipitation – currently around 2100 mm per annum, with consequences for the landscape as below. The Tywi river/lake at Dryslwyn after storm Gertrude last week, below. Several tried here – G. “Diggory”, G. “Clare Blakeway-Philips”, G .”Faringdon Double”, are lost without trace, or sulk and hardly increase in numbers -G. “Three Ships”, G. “Anglesey Abbey”. Such variety and fussiness inevitably present a great marketing opportunity for avid collectors and the supplying nurseries.

Thirdly, the origin of snowdrops in Britain is murky. Britain has a generally sparse native flora, compared with continental Europe, the Caucuses and Western Asia where most snowdrop species originate, following losses of much flora during the last Ice Age. Click here for some fascinating maps of where the ice coverage extended to, and how it retracted as the climate warmed. So, it’s imagined that snowdrops were brought (back?) into Britain, maybe by the Romans, maybe by the Normans. Wales had its own very special links both with Ireland to the west over thousands of years; with the Romans during their conquest in AD 43 -84 (click here) in Britain, although West Wales was inevitably one of the last areas to be reached; and then with Europe more generally through the monasteries, abbeys and the centre of mediaeval medicine associated with the local physicians of nearby Myddfai during the early middle ages. Click here, here, here and here for some examples.

Fourthly, from my observations of flowering snowdrops, there are real issues with pollination, and seed formation. Temperature and rainfall, certainly in West Wales present problems with the number of potential insect pollinators around at snowdrop flowering time. In our garden, early cultivars’ flowers often contain tiny flies, little larger than midges. (And now you will notice a trick gleaned from Naomi’s talk – a tiny Christmas Cracker mirror, about to be chucked out, and taped to a stick is really handy at looking up inside snowdrop flowers, to see if there are any insects present – without disturbing them). Are they there because the flower is a dry, and probably warmer, refuge? I’ve previously measured 2 degree C temperature differences within snowdrop flowers, compared to the outside ambient temperature. So they may well be thermogenic, as part of their strategy to attract early season insects for pollination purposes. I’ve also occasionally found Hebrew character moths, Orthosia gothica, resting at night  inside the flowers, later in the season.

If you’re lucky enough to have a nearby honeybee hive, then the bees love visiting the open flowers, and avidly collect the pollen. But at this time of the year, honeybees won’t be able, or willing, to fly long distances to seek out a snowdrop population. We haven’t seen any yet in the garden this year, but now have well over 100 types of snowdrops in flower.Finally, in our garden, overwintered bumblebee queens rarely emerge from hibernation before the middle, to end, of February. Too late for many snowdrops, but even with masses of Galanthus nivalis flowers available every year, I’ve only ever seen a single bumblebee bother to visit them, shown below. All this means that snowdrop colonies anywhere tend to be quite discrete, almost ‘island‘ communities, rarely running from one property to the next, in a way that many common ‘native‘ wild flowers do.

Fifthly, our part of Wales had large numbers of historic houses, many of which were built around the late 1700’s/early 1800’s. At the time, the most commonly available species of snowdrop in the UK was the familiar single ‘native’ Galanthus nivalis, (although there are some herbal records of G. plicatus being present from the 1600/1700’s), and the ‘native’ double form of G. nivalis flore pleno, which was first recorded in the UK in 1703. Since this seems sterile, at least as far as seed production is concerned, with vestigial ovaries, it could only have been spread by bulb movements. However, its double flowers are clearly also visited by the tiny flies, (3 below), so could act as a potential pollen source for pollinating single G. nivalis. But there is still huge variability possible within the single form of G. nivalis, judging by the number of available forms of it grown, and illustrated, on the ‘Judy’s Snowdrops’ pages. Click here for more. Many of the local grand historic houses would have had gardens planted up originally by the first generation of enthusiastic new occupants. They wouldn’t have had access to the more exotic species like G. elwesii, introduced much later towards the 1860’s/1870’s.

But huge numbers of these houses are no more. There is a wonderful website, a real labour of love, created by locally based photographer Paul White, who has spent over 25 years cataloguing some of these grand, and also not so grand Welsh houses, and their fall from grace, into dereliction. Paul won’t allow direct links onto his site (for security reasons, I guess), but do google “Paul White Welsh Ruins” and have a look. A couple of our now revered local gardens at Aberglasney, and indeed the NBGW fell into this category of property, but have been brought back from the brink, and have had wonderful gardens recreated.  No such luck, for example, at Derry Ormond just North of Lampeter, where the ornate gates, outbuildings and folly tower survive, but the huge main house was demolished in the 1950’s. Click here to see what it looked like. I have no idea whether snowdrops linger on within the grounds, out living the men, women and masonry of the 19th and early 20th centuries.

Sixthly, and finally, the ‘hunt’s’ aim is to visit some of these properties, and, if possible, with permission, to photograph and collect some small representative samples of their local ‘island’ snowdrops population, and grow them alongside our modest collection of named cultivars to see how they perform for vigour, flower production and fecundity – seed formation. In time I would love to be able to persuade an interested student to DNA profile some of these Welsh historic origin snowdrops, to work out whether there are any common genealogies, which would allow possible origins or sources of plants to be explored.

I have already made some enquiries, put out some feelers, and the first visits are planned. Readers may know that the NBGW was the first to DNA barcode the entire native Welsh flora. There are apparently 10 examples of individual G. nivalis snowdrops from around the world which have already been coded in this way, though frustratingly, I can’t find the specific Welsh record. What does a barcode look like though? This hidden code of life for the humble snowdrop? Courtesy of the excellent BOLD (Barcode Of Life Data system) website, the code for a single snowdrop is shown below – addendum -sadly I can no longer access this website. (It’s recorded as being supplied by one A.P. Davis, who I’m guessing may be Aaron Davis, one of the co-authors of the very detailed multi-author book on snowdrop identification – “Snowdrops – A Monograph of Cultivated Galanthus” 

In Letter code: (the letters stand for the 4 nucleic acids that form the basic building blocks of all genes – Cytosine, Adenine, Thymine and Guanine. Click here for more on the basic theories and limitations of DNA bar coding for life forms).

CTATTGACGCTCTCTGCCTTTTGCAATTTCATGTCTTGCGATTAGACATTAGGTATTGATGTGGAGATTGCCCCC CTCCTCATACATCATTAGGTGGTGGGTCTAATTGTTAGTTGTCGATCGTGATGGACGTGGCGAGAGGAGTTTTGA CACACATTCTATCATGGAAGTAACTTAGCTTGGGTGACGCACCGGAGTAACCTATGATAATTGGTGACATGTGAG TATTACTTGGAACATGACCCCAGGTCAGGTAGGGTCACCTGCTGAGTTTAAGCATATCA Illustrative Barcode:

Eventually, perhaps some of these local and historic forms of Welsh snowdrops may find wider use, as garden plants honed over the centuries to cope with the soils, climate, and rainfall of the Western fringes of the UK, and attractive to the few pollinators available in late Welsh winters. And will meantime grow comfortably in the environs of this once derelict, memory infused place, bringing their own magic with them.

Perhaps someone might even be able to study fungal root associations with snowdrop bulbs, since such symbiosis is likely to be critical to the success of this type of bulb in any given location. In the future, even nerdier galanthophiles may be able to scan bar codes of bulbs before shelling out hundreds on the latest new thing, or ‘breeders/nurserymen’ may even be able to predict, from reading the detailed barcode of a new seedling, just what it might look like, way before it’s grown to flowering potential – it can take 5 years from seed to new flower.

Probably all fantasy, and certainly for the average gardener like me, who simply wants bulbs that grow, thrive and delight at this grey time of year, an irrelevance.

The hunt will no doubt last many years, but in time I hope I can provide a few updates as to how it’s progressing. I am already very grateful to the people who have so far agreed to help me in this task, and been very generous with their time and advice. And if any readers know of any local ‘naturalised’ populations which might fit into the categories of being ancient, vigorous and floriferous, and which they think should be included in the project, do get in touch.

At the time of writing this, February1st/2nd, we’re at the date of the old Celtic festival of spring – Imbolc, or as it was known in Wales, Gŵyl Fair y Canhwyllau. Click here for more. The day that it was deemed to be light enough to dispense with candles for illuminating the livestock, at feeding time. How they coped without electricity is beyond me – we’ve resorted to my old SAD light box being switched on at mealtimes, after this gloomiest of winters, to help to reduce the melatonin influence, and aid sleep. It seems to be working, just as Galanthus ‘Imbolc’, a snowdrop originally from Oxfordshire, emerges spot on time, for once.And remarkably, from the still soggy garden, a few pictures of other early flowers, prompted by the mild conditions to peek out, grow fast, and then flop over in the rain.The above image shows the extremely odd mix of Crocus flowers fully opened under such poor light, but mild, conditions typical of early 2016. the light was so poor that the exposure was at  maximum aperture of f1.8, and a thirtieth of a second at 200 ASA, for the photographers out there.