Outside many homes on Bermuda, curious-looking bottles hang from porches. These are shark oil barometers, homemade devices that for nearly 300 years have been used by islanders who swear that the oil can predict the severity of approaching storms and hurricanes.
Fishermen contrive the oil by extracting it from shark livers, filtering it through cheesecloth, then sealing it in bottles. Folklore holds that when a storm approaches, the clear oil turns cloudy or spins at the base of the bottle, “spiraling like the tempest to come,” as described by one of many local news articles on the barometers.
While the oil has long been a source of local pride, naysayers scoff at its reliability as a forecasting tool. Little is known about the origin of shark oil barometers and no one can say exactly what causes their contents to shift from clear to cloudy when the weather turns. This summer, BIOS scientist Mark Guishard and college intern Shane Antonition paired up to begin looking at the mysterious little bottles through a scientific lens to find out just how useful these oil-filled tools are at forecasting storms.
“There’s no question that the oil reacts to changing weather,” said Guishard, a meteorologist and former director of the Bermuda Weather Service who currently leads BIOS’s Risk Prediction Initiative. “We want to find out, is there something to the patterns we see? Or is it like trying to read tea leaves?”
Traditional barometers were developed in Italy during the 17th century, and since then have been used by mariners, farmers, explorers, field researchers, meteorologists, and ordinary citizens to interpret approaching weather conditions. Originally filled with water or mercury, they were eventually replaced by mechanical versions. Barometers work by measuring changes in air pressure. This information is revealed on the instrument’s dial, a digital display, or today even through applications on smart phones and computers equipped with barometric pressure sensors.
Keeping track of changes helps to forecast the weather for the upcoming 12 to 24 hours. High air pressure generally brings fair weather. Low, or dropping pressure, indicates the approach of windy, stormy, or wet weather.
In contrast, shark oil barometers provide information in a decidedly less scientific and more subjective fashion. Clear oil means fair weather. Cloudy or shifting oil means turbulent weather. One theory about these changes suggests that shifts in the atmosphere affects the oil in a live shark, which then acts as an alert for the animal to move to deeper water before a storm. The oil may retain this function when removed and the change remains visible, the oil transitioning from a clear golden color, like cooking oil, to milky white.
But very little research has been done into the oil’s composition, origins, and historic use, which users say that to be accurate must be taken from a shark during certain conditions, such as during a full moon, and specifically from a dusky (or puppy) shark. Historians say that Bermuda may stand alone in its use of shark oil in barometers. Jane Downing, a registrar at the National Museum of Bermuda, said shark oil may be a variation on bear grease barometers used by native islanders, especially by those living on St. David’s.
Sharks parts, from fins to internal organs, have been used for a variety of reasons through the ages, including for culinary, cosmetic, and medicinal purposes. “It’s possible that when people began draining shark livers of oil, they simply began noticing that the oil reacted to weather changes,” Guishard said.
Antonition and Guishard approached their topic during an eight-week summer project sponsored by the Bermuda Program at BIOS. With only a short window to work before Antonition needed to return to studies at Plymouth University in England, they focused on developing methodologies for monitoring and analyzing the consistency of the barometers’ oil. Methodically capturing the changing appearance of the oil is enabling the first rigorous comparison between the device and detailed weather data provided by the Bermuda Weather Service (including information about humidity, sea level pressure, temperature, and dew point).
Antonition conducted one experiment to monitor changes in the oil as weather patterns shifted by setting up a time-lapse camera to snap two barometers at 10-minute intervals for one week. He then analyzed the specific behaviors of the shark oil by characterizing the opacity, from clear to cloudy, and comparing the information he gathered to data presented by the weather service.
“The time lapse was interesting for two reasons,” he said. First, the data demonstrated a potential link between changes in weather and the changes in the oil’s composition, something he hopes to better define. Second, it seemed to indicate a valid methodology.
“It was a challenge to develop a system to continuously monitor and analyze the behavior of our shark oil barometers over an extended time period that put the subjective nature of reading the devices under an objective scientific lens,” he said. “In the end, it is not only a methodology we can be proud of, but is one we can develop further and improve using what we’ve learned from the project.”
Antonition also conducted lab experiments with two bottles of shark oil to look at pressure (which would help when looking specifically at how the oil would react during a hurricane) as well as temperature (at a range of degrees).
“The results of the pressure experiment were interesting for the lack of any correlation in the lab,” he said. While the time-lapse experiment hinted at some influence of pressure on the shark oil, no changes in the oil were observed when pressure was the sole variable manipulated in the lab experiment.
Since the devices are called barometers, devices that by definition are expected to measure changes in atmospheric pressure, “to find no apparent link between pressure and the behavior of the shark oil was quite intriguing,” he said, adding that future experiments are needed to validate the findings.
Temperature experiments provided insights as well. “It was fascinating how responsive the oil was to temperature,” he said. As temperature increased, the oil cleared. When temperature decreased, the oil clouded. “While the experiment in itself does not directly tell us about weather forecasting, I feel it is still critical for identifying the mechanisms that influence the behavior of shark oil.
One downside of the timing of their summer project was that no major storms or weather events hit the island, making it difficult to actually monitor behavior in the oil through a variety of weather events. Among the next steps listed for the project are longer time spans for monitoring the oil under different atmospheric and weather conditions, as well as different seasons; expanded pressure experiments; and chemical testing of the oil, to learn more about its specific behaviors and properties.
Antonition said he’d also like to see a more complete literature review on the use of shark oil during centuries past.
Is he a believer in shark oil barometers? The verdict is still out. “There’s some validity to them,” Antonition said, adding that while he grew up with one hanging at his childhood home he doesn’t view them as “entirely accurate.” He also said that he is open to being proven wrong, and hopes to continue his work at BIOS in years to come for more conclusive answers.
But whether skeptic or steadfast believer in the oil’s meteorological properties, he said they are worth studying simply because they are “a rich piece of our island’s maritime cultural heritage.”
Sponsored story by Bermuda Institute of Ocean Sciences.