97 
Leaving Matanzas on March 25, a two hours’ wait for a train 
enabled us to examine the flora of a coral-limestone hill at Em- 
palme, where specimens of twenty-five species not seen in the 
Matanzas region were secured. Proceeding the same morning 
to Madruga, in Havana province, that town was made a collect- 
ing base for four days. It is situated among hills, nearly on the 
backbone of the island, the underlying rock being an eruptive 
over an area of several square miles, apparently surrounded on 
all sides by the coral-limestones. Sulphur springs, popular for 
baths, cause Madruga to be visited by Cubans. The flora of this 
Fic. 13. Hut, thatched with palm leaves. Coconut palm near center of pho- 
tograph. 
area of eruptive rocks proved to be almost wholly different from 
that of the coral-limestone soil. In fact, along the contact of 
the two formations south of Madruga the change in the vegeta- 
tion is so abrupt as to enable one to determine from a glance at 
the plants whether he is over limestone or traprock, and we 
located the contact lime within narrow limits for several hundred 
yards. Specimens of 185 species were obtained at Madruga, 
nearly all different from those previously observed by us. 
