IN 1674 George Ravenscroft, an English glass merchant, was granted a patent for the discovery made at his factory in London that adding lead oxide to the melt resulted in a clearer, more durable product. Thus was born lead crystal, and with it the fashion, in England, of drinking wine from glass vessels rather than, say, pewter ones.

Wine glasses have evolved since then, of course, and one aspect of this evolution is of particular interest to Theresa Marteau and her colleagues in the Behaviour and Health Research Unit at Cambridge University. Dr Marteau suspected that glasses have got bigger over the years, and that this may have contributed to the increased drinking of wine in Britain—an increase that has been particularly marked in recent decades.

As they report in the BMJ, a medical journal, she and her team obtained data on glass volumes going back to about 1700 from sources including the Royal Household (which buys a new set for each monarch) and the Ashmolean, the university museum of Cambridge’s arch-rival, Oxford. Altogether, they recorded the capacity of 411 glasses and, as the chart shows, there has indeed been a near-continuous tendency for that capacity to increase since Ravenscroft’s day (he died in 1683). There is also a notable acceleration of the process starting in about 1990. In all, the average capacity of a wine glass increased from 66ml in the 1700s to almost 450ml in 2016-17.

 That this volumetric inflation has stimulated wine consumption—Dr Marteau’s second hypothesis—is hard to prove. But it may have done. The amount of wine drunk in Britain has risen more than sevenfold since 1960, while the population has grown by only 25%. Data collected between 1978 and 2005 by Britain’s Office of National Statistics suggest the proportion of adults drinking wine fell from 60% to 50% over that period, while the average weekly wine consumption of those who did drink the stuff tripled, when measured as units of alcohol. Another set of data, collected by the Institute of Alcohol Studies, a charity, suggest that the amount of alcohol from all sources (measured as pure ethanol) consumed per head in Britain is about the same as it was in 1980, though it has fluctuated quite a bit in the intervening years, peaking in 2004.

Meanwhile, work designed to test directly the idea that glass size matters, which Dr Marteau published last year, produced mixed results. She looked at the consequences for wine sales at a bar in Cambridge of serving its wares in both bigger and smaller glasses than normal, while keeping the serving sizes on offer (125ml or 175ml, according to customer choice) the same. In weeks when the bigger glasses were used, wine sales went up by 9% on average. The larger vessels, it seemed, were indeed encouraging customers to order refills more often. On the other hand, in weeks when the size of the glasses was below normal, sales did not go down. Reducing glass sizes, then, does not keep people sober.

Correction (December 20th): A previous version of this article described the Institute of Alcohol Studies as a temperance charity. This has been changed.