80 MASS. EXPERIMENT STATION BULLETIN 402 



The Experiment in Forecasting 



Beginning in 1923, following the publication of the first discussion of the ap- 

 parent relation between weather and keeping quality, the writer undertook to 

 make experimental forecasts of keeping quality in the Wareham- Carver area in 

 Massachusetts. The reasons for choosing this region are obvious. This was the 

 area in which Mr. Griffiths' inspections had been made and to which his estimates 

 would most nearly apply, and the presence of the cranberry station made avail- 

 able weather records taken with special reference to cranberry growing. 



At the outset, it was determined to publish the forecasts, not because of any 

 practical value which the}' might have but because of general interest in the 

 subject and to avoid any suggestion of partiality to special interests. From 1923 

 to 1928 the forecasts were published in the Wareham Courier, the official organ 

 of the Cape Cod Cranberry Growers' Association. At first, a general forecast 

 was made between September 10 and 15; but beginning in 1927, separate fore- 

 casts were issued for the early and late berries, the first between September 10 

 and 15 and the second between October IC and 15. 



With a single exception, these forecasts were substantiated by the subsequent 

 behavior of the berries. This exception was an unusual amount of rot which 

 occurred in the Howes of this region in the crop of 1926. The poor keeping quality 

 of the Howes in this year was due, in the opinion of many competent observers, 

 to the fact that in 1926 the Howes were picked unusually green and it is well 

 recognized that this is unfavorable to best results with this variety. Even aside 

 from any such explanation, the record is far better than could have been expected 

 in view of the difificulty and the uncertainty involved where so many only partly 

 understood factors are operative. 



To a certain extent the publication of these "forecasts" tended to obscure the 

 results of the experiment. Many cranberry growers, as well as inspectors and 

 sales agents, were much interested in the work and kept in close touch with it 

 and their actions were to some extent influenced by the published forecasts. For 

 example, it was inevitable that, after the forecast of relatively poor keeping quality 

 made in the fall of 1929, there should be unusual care exercised on the part of 

 inspectors and the most interested shippers in Plymouth County. Several lots of 

 berries known to be of poor keeping quality were sent to the cannery at once and 

 others were handled by consignment to nearby markets. This was, of course, 

 highly desirable but to a certain extent it obscured the accuracy of the results. 



On the basis of the experience of these thirteen years, one might well con- 

 clude that it is possible to forecast with a fair degree of accuracy the keeping 

 quality of the cranberry crop of Plj^mouth County from a study of the weather 

 alone; but in practice it would no doubt always be safer to check this by actual 

 tests of sample lots of the berries themselves; not only for the purpose of having 

 an additional check on the crop as a whole but because this would make it pos- 

 sible to determine the variation in the keeping quality of berries from individual 

 bogs. Because of the close relation of the two subjects, a discussion of incubator 

 tests is included in the present paper. 



Incubator Tests of Keeping Quality 



In his annual report as chairman of the Board of Inspectors for the New Eng- 

 land Cranberry Sales Company, 1922, in discussing the problem of distributing 

 berries in such a way that rejections would be reduced to a minimum, Mr. 

 Griffith said: "I believe the ultimate solution of this problem rests in the develop- 

 ment of some method of testing berries by which we may determine their keeping 

 qualities." 



