William Wordsworth once wrote that "nothing can bring back the hour of splendour in the grass, of glory in the flower." But curators at Harvard's Botanical Museum are trying to do just that, as they embark on an unprecedented effort to restore the museum's renowned collection of glass flowers. Commissioned more than a century ago as teaching aids for Harvard's botany classes, the flowers are among the finest glass objects ever made— some 4,000 individual life-sized pieces representing more than 830 plant species. The replicas have a fresh-picked realism that makes most visitors do a double take.
A glass botanical model of diseased strawberry plants has acquired an age-related alkaline film.Photo by Grant Delin
But the glory in these flowers is starting to fade. Networks of tiny fissures have marred the beauty of a sage blossom. The painted coating on a piece of glass seaweed is peeling away from the underlying form. The leaves of a pear tree show blotches of a frosty white film. Cactus spines have spontaneously snapped.
"It's like glass disease," says Susan Rossi-Wilcox, the curator in charge of the restoration. Although much of the damage is normal aging, she says, plans to repair it are complicated by the very factors that bring the glass flowers to life: fragility, complexity, and idiosyncrasy. "Artists still look at certain features of the collection and tell me they can't figure out how they could have been made," she says.
That's the trouble. The flowers are remarkably precise renderings of real plants, accurate to the most minute detail. They resemble no other glass objects on Earth: No vase, no goblet, no stained-glass window has such an elaborate profusion of forms. The needles on a pine branch, the seeds on a strawberry, the hairs on the back of a pollinating bee— all were shaped, colored, and attached one at a time, by hand. Most pieces have hundreds of individual parts, each made of a different glass with unique properties. And in order to repair them, one needs to know something about how they were put together in the first place.
"The most unusual problem with these particular artifacts is that some are composites of different materials," says Carlo Pantano, a professor of materials science and engineering at Pennsylvania State University who has performed microscopic and X-ray analyses of the flowers to determine their structure and composition. In addition to glue, paint, primers, and glass, he says, some have wire skeletons and a varnish that imparts a matte finish. And no two seem to be constructed in the same way. "That's what makes them so realistic," says Pantano. "If these things had been pumped out by a machine, they'd be perfect, they'd never break, and they'd look totally fake."
Instead, each one was crafted by a father-and-son team of German glass-workers hired by the first director of Harvard's Botanical Museum, George Goodale. When Goodale approached Leopold and Rudolf Blaschka in 1886, they had already earned an international reputation for their intricate replicas of anemones, jellyfish, and other marine creatures. Eager to improve on crudely rendered wax or papier-måché botanical models, Goodale persuaded the Blaschkas to try their hand at the plant kingdom, and by 1890, the craftsmen were working on the flowers full-time. After his father died in 1895, Rudolf continued alone, sending carefully packed wooden crates by steamer to Boston twice a year until his retirement in 1936.
Unfortunately, no one's sure how the Blaschkas plied their craft. Although they began with standard practices and materials that had sustained their mass production of marine models, it's clear that after his father's death, Rudolf became a glass virtuoso, improvising techniques and materials with each new creation. With his father, for example, he had shaped clear forms from commercial glass and painted them. But by 1900 he had stopped using paint altogether and began tinting his own glasses. He would fuse layers of tinted glass to create brilliantly colored, long-lasting enamel surfaces.
In X-ray studies, Pantano has found that the enamels were doped with chromium, copper, and other elements to impart color. He also found that, over the years, ions in the glass reacted with water molecules in the air, leaving a hazy white alkaline film on the surface of many replicas. Corrosion also weakened the glass by pulling atoms within the models to the surface. The migration leaves microscopic pores in the glass that can result in networks of tiny cracks called crazing.
The addition of doping agents for color would also have altered the rate at which the glass contracted as it cooled, creating stress between parts with different colors and compositions. The repeated heating and slow cooling required to fuse many parts or many layers of glass enamel would also have weakened the structure of the models, says Pantano.
The collection includes a remarkable glass rendering of a bee pollinating a milkweed flower.Photo by Grant Delin
"There's a tremendous amount of stress built into some of the parts," says Pantano. "The more you handle them, the more chance of breakage— which may be why nobody's even dusted them off in decades."
Cleaning has to be the first stage in the restoration, says Pantano, because dirt will foil any attempts at gluing or patching. But a feather-duster approach can't dispatch a century of particulates, and solvents might degrade the organic glues and primers used in earlier models. Instead, Pantano has proposed bathing the flowers in a charged stream of oxygen gas that combines with the organic matter in dust to produce carbon dioxide and steam. The treatment would eliminate offending particles without physical contact. (See also "Rocket Science and Art Restoration," page 70.)
Removing the alkaline film, in contrast, will probably require a more aggressive technique not unlike sandblasting, albeit with particles of dry ice rather than sand. Abrasive jets of frozen carbon dioxide are already used in the semiconductor industry to clean computer chips. If that doesn't work, Pantano says, he'll explore burning off the corrosion with lasers.
Finally, Pantano hopes to use a liquid glass called a solgel to mend the models. The substance is an alcohol-based silica compound that dries and contracts as the alcohol evaporates, leaving a solid adhesive whose molecular structure is almost identical to glass. A high-tech polymer glue with optical properties similar to those of glass would do the trick, too, he says. "But we're purists, and the idea of fixing glass flowers with plastic just isn't appealing." Fusing more glass onto breaks is out of the question. "If you put more heat to these things, they're just going to shatter."
Museum officials project that the restoration could take at least five years and cost more than $2 million. But the most vociferous objections to the project have nothing to do with Wilcox. More people are dismayed that the flowers will have to be taken off display while undergoing repair. Once the models are restored, curators hope to acquire modern display cases to replace the antique oak ones that have plenty of leaks as well as plenty of character. More tightly sealed cases will help protect the flowers against dust and humidity— the main threats to their splendor.
"One of the reasons Goodale wanted them made out of glass was that he believed they would be forever," says Pantano. "And he was almost right."
In 1982, Harvard University published a small book filled with breathtakingly beautiful photographs and botanical descriptions of the glass flower collection. The book, The Glass Flowers at Harvard, is available by special order from bookstores or from the museum shop at the Harvard Museum of Natural History. To special order from the museum, visit the shopping page at their Web site (www.hmnh.harvard.edu) for an order form.
The Web site of the Harvard Museum of Natural History has a brief overview of the history of the glass flowers and information about the ongoing exhibition "Modeling Nature": www.hmnh.harvard.edu/exhibitions/glass.html.
Visit the Web site of the Corning Museum of Glass for more information about glass manufacture: www.cmog.org.