XVI 



It has been deemed preferable, however, all other things being 

 equal, to refer to some readily accessible and popular work. But 

 in cases where such works do not give the limits to the families 

 which have been indicated by the most approved researches, 

 references are made to the monographs or other publications 

 wherein the information is furnished. Some of the families, 

 however, have not yet been assigned the limits which, in the 

 opinion of the compiler, appear the most natural ; in order, 

 therefore, to indicate as nearly as possible the relative values of 

 the respective groups, the system of notation recommended espe- 

 cially by the late Hugh Strickland has been adopted. When 

 there is an exact ecjuivalency, either as to the limits assigned by 

 the diagnosis, or as to the contents, the sign of equality ( = ) is 

 used ; when the group referred to is larger than that adopted, 

 the corresponding sign (<) is prefixed to the former; when the 

 group referred to is smaller, the usual sign indicative thereof 

 ( > ) is prefixed ; and when the group referred to is entirely 

 different, including some forms not in and excluding others re- 

 tained in the group compared with it, the sign ( x ) is employed 

 as a prefix. 



ACKNOWLEDGMENTS. 



In the appended list of authorities, and in connection with the 

 names of the families, will be found the references to those 

 authors who have been followed in especial cases. The compiler 

 would also especially acknowledge his obligations to Mr. W. H. 

 Dall for various kind offices and assistance in the preparation of 

 this list. 



