Abstract
LONDON.
Challenger Society, June 27.Dr. E. j. Allen in the chair.Dr G. H. Fowler: A statistical method of analysis of tidal stream cbservations. By plotting on a diagram of 360Ã the exact direction of the tidal stream at every hour during a complete lunation, the profound effect of prolonged wind and other extrinsic causes became obvious. Hence in order to obtain a normal (probable) result, abnormal observations must be neglected. This is most safely done by arranging all observations at each hour under their proper directions, plotting them into curves of frequency, and rejecting those observations which are then seen to be due to premature or belated turns of the stream, etc. As the vailahle observations were not numerous enough for other methods, they were grouped under sixteen points of the compass, for example, all observations between N. 12 XV. and N. W. were grouped as N.N.W. For each hour after H.W. Dover, the value of each compass-point in degrees of a circle was then multiplied by the number of selected occurrences; the sum of the pro- ducts divided by the sum of the factors then gave the probable direction of the stream at that hour. Velocities were then simply meaned; and from the data thus obtained ellipses were constructed which showed the direction and velocity of the stream at each hour after H.W. Dover for twelve hours.C. Tate Regan: The distribution of the Clupeime. The anadromous habit of the fishes of the shad group leads to localisation and the evolution of genera and species with a restricted distribution; in the strictly marine herrings, sardines, etc., with pelagic. larvte, the majority of the genera and species are more widely distributed.
Article PDF
Rights and permissions
About this article
Cite this article
Societies and Academies . Nature 99, 398–400 (1917). https://doi.org/10.1038/099398a0
Issue Date:
DOI: https://doi.org/10.1038/099398a0