36 The Philippine Journal of Science ish 
approximately be assigned to the first or second centuries of our 
era. At any rate the absence of Sanskrit roots in the Polynesian 
languages indicates that the Polynesians were never in direct or 
indirect contact with Sanskrit peoples. It is apparently quite 
safe to assume that the Polynesians have occupied the islands of 
the Pacific for at least two thousand years, perhaps longer. 
This period is sufficiently long for man to have wrought great 
changes in the character of the vegetation of the different islands, 
to have destroyed the original forest over large areas, and to 
have provided the proper habitats for the light-loving weed-flora. 
Safford’s work on the flora of Guam is an alphabetic arrange- 
ment of all the species known to him from that island. There 
is no summary by families and genera, so that it is rather diffi- 
cult to gain an adequate idea of the constituent species of the 
flora from an examination of his work. In nomenclature Saf- 
ford’s work follows the American Code which is based on strict 
priority and admits of no generic list of nomina conservanda. 
The idea of generic types was also applied with some surprising 
results, and, in at least some cases, the type adopted does not 
appear to me to be the logical one. For the nomenclature 
adopted in Safford’s work Mr. W. F. Wight is primarily respon- 
sible, and is the authority for most of the changes of names 
included. In this rather popular work, and one that is strictly 
alphabetic, it is rather curious to note not only new combinations 
and new specific names, but also new generic names and even one 
new family name. It hardly seems probable, especially in view 
of the fact that there were available three properly constructed 
family names already, Lamiaceae (1836), Nepetaceae (1843), 
and Salviaceae (1879), that the new name, Menthaceae, proposed 
for the Labiatae, will meet with general acceptance. The pro- 
posal of the name Menthaceae, under these circumstances, appears 
to me to be entirely unwarranted, and quite at variance with the 
principle of priority. 
In the following enumeration the nomenclature has been 
worked out on the basis of the International Code of Botanical 
Nomenclature, and accordingly the accepted generic and specific 
names will in some instances be found at variance with those 
used by Mr. Wight in Safford’s work. The author is responsible 
for the identifications of most of the specimens cited in the 
present paper, the material examined, as noted above, being the 
collections of Mrs. Clemens, Mr. Thompson, Mr. McGregor, and 
a native collector working under Mr. Thompson’s direction, a 
total of 824 numbers. For purposes of comparison I have had 
