which by 0 6 0 0 ~ ~ ~ had backed to 150 degrees. This would increase the directional shear between the surface and the air at higher levels if Valentia was representative of those conditions. With regard to wind speed, again there was a noticeable wind shear at low levels. Between 0000 and 0600 GMT the surface wind at Roches Point varied between 11 and 17kn compared with 30 and 40kn at 1286 and 2813m respectively during the Valentia ascent. Such vertical profiles are by no means unusual, but they were clearly sufficient to provide an environment favourable for tornado development. Data for Britain show that the months with the highest frequency of tornadoes are June, July and August, while February, March and April have far fewer tornadoes than other months. With regard to the time of day, 70 per cent of all recorded tornadoes occurred between 1200 and 1 8 0 0 ~ ~ 7 ' (Meaden 1976). These patterns reflect the r d e of diurnal and seasonal heating in creating instability. The Youghal tornado falls outside these preferred seasons and times. However, the environmental conditions associated with the Youghal tornado do correspond with the second most common set of environmental circumstances which produce tornadoes in the British Isles, when cold air overlies warm seas. The atmospheric conditions that spawned the tornado at Youghal are not unusual for Ireland. Local guardai (police) report that a similar event occurred on the east side of the town two years earlier. Elsewhere in Ireland there have been recent reports of other tornadoes, during March 1995 and subsequently in Coachford, County Cork (Lennon 1995). However, their extremely local nature and the sparsity of population in many regions means that they are likely to be significantly underreported. The low recorded frequencies of tornadoes in Ireland are unlikely to be due to unfavourable environmental conditions for their development.