Forecasting the weather on the west coast of Africa is made difficult by the inaccurate pressure reports of ships at sea and of land stations whose elevation are not definitely known. However, from reliable reports of coastal stations a relation between pressure changes and wind variations is demonstrated, (see fig. 2–4). In each example a 24–48 hour pressure fall with onshore winds all along the coast was followed by a 24–48 hour pressure rise with off-shore winds. Two examples of pressure variations on a ship's barogram when a “tornado” occurred are given in fig. 5–6. Each diagram shows a falling pressure tendency followed by a rising tendency, with the “tornado” at the minimum pressure in the first case and 10 hours after the minimum in the second case. Evidently whenever these tendencies are observed together thunderstorms should be forecasted. This is born out for the statistics July to October 1934 which show 15 thunderstorms on the 22 days when the pressure was rapidly rising, and only 10 thunderstorms for the other 61 days!
*Translated with some abbreviation from Annalen der Hydrographie, v. 64, 1936, pp. 107–111.—Edna S. Stone.