149 
this phenomenon, on any other plant but the Frostweed, Heli- 
anthemum canadense, which receives its popular name from the 
occurrence. A definite and complete list of plants displaying 
such crystals would be interesting and might throw more light 
on the physical reasons therefor. 
BOOK REVIEWS 
Johnston’s monograph of the genus Cordia in 
Southeastern South America 
Under the modest title ‘‘Observations,’’ this accomplished 
specialist in the family Boraginaceae contributes a paper (Con- 
trib. Gray Herbarium X CII) which well sustains the reputation 
of that institution for comprehensive and accurate work. The 
keys to the several sections of the genus, the citations of publi- 
cation and synonomy and the descriptions of new species are all 
excellent, as well as the citation of the specimen examined, but 
the most striking and valuable part of the work is in the ex- 
tended discussions of the composition of the species. It is a la- 
mentable fact that the taxonimist is usually unable to reach 
very positive conclusion as to identity from the majority of 
descriptions, without referring to type specimens, but it would 
be rather difficult for any careful analyst to be in doubt regard- 
ing the identity of his specimen after reading Dr. Johnston’s 
discussion of its essential characters and variations. 
Every student of tropical American plants must have been 
puzzled in his attempts to identify specimens of the difficult 
genus Cordia, even when he has good material for comparison. 
Dr. Johnston’s ‘“‘observations” regarding variations in the num- 
erous specimens examined should contribute much toward ac- 
curacy in future publications of this genus. We cannot help 
regretting that the author should take so comprehensive a view 
of the limits of this genus. To the present writer, it seems that 
the generic distinction between groups which Dr. Johnston re- 
gards as sub-genera are greater than those which separate 
Cordia from its relatives. Both accuracy and convenience would 
seem to dictate the separation of Cordia, as here treated, intoa 
number of genera. 
H. H. RusBy 
