Rainfall Correlations in Trinidad

Abstract

IN connexion with my communication to NATURE (February 7, p. 192) on the above subject, it may be of interest to mention that I have, recently received a letter from my friend, Dr. Preston E. James, Department of Geography, University of Michigan, informing me that at the meeting of the Association of American Geographers at Washington last December, he recorded in an address on “Geographical Factors in the Trinidad Coconut Industry” certain correlations between rainfall and coconut yield. In connexion with data obtained from a large estate in the extreme south-east of the island, he found a positive correlation between the rainfall of one six-month period and the quality of the nuts six months later, “quality” being a matter of the proportion of selected nuts—which will not pass through an iron ring 4 inches in diameter—and “rejects” and “culls”—or those that will. (This grading into “selects” and “culls” is the recognised commercial practice where coconuts, as nuts and not as copra, are exported to northern markets.) Dr. James claims to have found a positive correlation of 0.733 ± 0.072. The correlation for the same six-month period, that is, without any lag, was 0.508 ± 0.109.