PREFACE 



IlIl'XIIiKT that, thr second part of thcs< Tulili \ /<// & 

 metricians has hern so long in making its appraianr. . S. \. ial l-i-toi have 

 combined to cause delay: first and most important IMS hrm an iindi-r-estiraafce 

 on my part of the time and ease with which certain tables could be computed, 

 tables that I felt were desirable; and in the second place it was not financially 

 possible to publish them without the aid provided by their being issued first in 

 Biometrika. The majority of the present tables are printed from stereos moulded 

 from type originally set up for that Journal. If subscribers to Bionut 

 raise perhaps not unreasonable objection to the early reprint of some of these 

 tables, they may be reminded that others have been in use for many years, and 

 that this use would have been impossible, had they been withheld till they could 

 appear with a completed collection. The first series was published seventeen years 

 ago, and the rule has been to publish each table as soon as it was computed. This 

 second series contains the major portion of the computing work carried out by 

 staff and post-graduates in the Biometric Laboratory during the past seventeen 

 years. I say 'the major portion' because there are two important exceptions, which 

 are closely associated with these Tables for Statisticians and Biometricians, but the 

 great extent of which prevented their inclusion in any book like the present. I 

 refer to the Tables of the Incomplete Y '-Function and to the Tables of the Incomplete 

 B-Function. The former has been published as a separate volume of the same 

 format as the present by H.M. Stationery Office and the latter is just completed 

 and ready for printing*. In addition there are of course the tables issued in the 

 Tracts for Computers series, but these largely appeal to mathematicians as well as to 

 statisticians, and to have reproduced certain of these tables here would have 

 much overweighted an already ample volume. 



One further remark may be made ; there has been a considerable demand for 

 those parts of Biometrika in which these tables were originally published, and it 

 has not been possible to meet fully this demand, as doing so would have rendered it 

 impossible, without much reprinting, to provide complete sets of that Journal to 

 many libraries desiring them. The present issue will enable the Bionu-tric 

 Laboratory to supply students and others with the tables they need without 

 impairing the small remaining stock of sets of Biometrika. It will also be a relief 

 to those librarians, who find that the volumes of that Journal containing tables are 



* The series ought to include Table- of the Incomplete G-Function, Le. the Probability-Integral of 

 the Type IV curve, but the age of the present Editor is likely to preclude his superintending any task, 

 which even exceeds in the magnitude of its calculations that of the Incomplete B-Function. 



