370 BOTANICAL GAZETTE [DECEMBER 
coccids, and other insects are closely associated, is the identifica- 
tion of the organisms concerned in the complex. Of course, the 
principle of significant figures must be considered in biology, as in 
physics and mathematics. The most detailed field observations 
and carefully planned experiments may lose much of their signifi- 
cance if an investigator fails to secure adequate information con- 
cerning the identity of the plants and animals with which he is 
working. In most cases, therefore, it is advisable to prepare 
museum specimens which may be preserved as a record for verifica- 
tion by other investigators. Typical herbarium specimens of all 
of the plants to be discussed in this and in subsequent papers have 
been deposited in the Gray Herbarium of Harvard University. 
Specimens of the insects have been preserved in the collections of 
Professor W. M. WHEELER, who collaborated with the writer in 
the investigation of the myrmecophytes of the Kartabo region. 
The ants were identified by him, the coccids by Mr. HAROLD 
Morrison of the United States Bureau of Entomology, and certain 
parasitic Hymenoptera by Professor C. T. BRUEs. 
In dealing with Cecropia, one is concerned with a group of plants 
which present many taxonomic difficultiés. The leaves of mature 
plants frequently are much too large for herbarium sheets of 
standard dimensions, and the inflorescences are difficult to season 
for museum purposes. Thus many of the descriptions of species 
are based upon the study of more or less fragmentary ‘material. 
Furthermore, there appears to be as yet no consensus of opinion 
as to which of the foliar and floral characters are of the greatest 
diagnostic value. In view of these facts, the writer devoted 
considerable attention to the investigation of the varia- 
bility of the morphological characters of the Kartabo species of 
Cecropia. 
There proved to be two distinct species, a myrmecophytic 
species, with a well developed trichilium and numerous food bodies, 
and a non-myrmecophytic species, which is entirely devoid of 
these structures. The latter is considered by Dr. E. H. SNETHLAGE 
to be a variety (decurrens) of C. sciadophylla Mart. The former 
does not agree with any previously described species, and owing 
to its association with ants, was studied more intensively. 
