Kleinere Mitteilungen. 



Further researches on some statistics of Coffea 



(fourth comuiuuication). 



By P. C. van der Wölk, Buitenzorg-Java. 



(Eingegangen am 16. Febniar 1914.) 



The investigations which now follow are in close connection with the 

 three former communications concerning the same subject'). In the course 

 of time these investigations have indeed assumed a very special character. 

 They were begun with such a manifestation of a vast array of numbers 

 obtained by the measurement of different varieties of Coffea which has 

 furnished a proof by means of statistics of the existence of multiple factors 

 for a single external property. Though my further investigations have at 

 the present time yielded also other results, so the above mentioned have 

 in consequence become the guiding line of the publications, and so I have 

 as it were concentrated my attention on this single point vnth regard to 

 still further working up this question. In this way it has indeed appeared 

 that the statistics are easier to handle than in many instances one would 

 well suppose, so that deeper conclusions may indeed be drawn than many 

 people would readily credit. In the present time, one sees in the "stati- 

 stics" more properly a control system than a basis for progressive research. 

 What biometricians have extracted from their ciphers for the benefit of 

 progressive ideas, has, in general, been somewhat sharply criticised up to the 

 present time. 



Since Johannsen, much life which may take shelter in statistics is set 

 forth completely. 



But this sceptical and very critical standpoint has, without doubt, been 

 very useful for the solid development of statistics. Indeed, we have to be 

 very prudent with the so-called statistic method and it must be stated that 

 in the name of statistics many follies have been done. 



However, we have no occasion to gaze blind at the sterility of the 

 statistics. The present statistics do not see the genesis of a table: they 



') See Vols 10 and 11 of this periodical. 



