A NATURALIST IN BRAZIL 



butterfly with yellow crossbands, for the coconut-palm. The Thoas, 

 a great black and yellow Swallowtail, breeds on the lemon-trees, 

 and the brown and red, gently fluttering Danaid on the poisonous 

 Asclepiadiae, to which family the broad-leaved Wax-plant belongs. 



For even the poisonous plants must be utilized, and they have 

 their lovers, whose digestive systems are so devised as to tolerate 

 their poison. Similarly, the venomous snakes are devoured by the 

 handsome black Mussurana, a harmless snake, and the Skunk kills 

 the largest rattlesnake without heeding its fangs. 



What I have said of the butterflies is true likewise of the beetles, 

 grasshoppers and bugs. If an insect is adapted to feed on a certain 

 plant, it will inevitably find its way to its pasture, since it is only 

 on this plant that the mother insect will lay the egg from which the 

 insect is developed. The insects which grow to maturity on plants 

 of other species cannot deprive them of their food, since for them 

 it is not food at all. There are, of course, "polyphagous" creatures, 

 which can live on several kinds of plants. We might explain their 

 evolution as follows : At first each plant would have supported one 

 species of insect. After repeated experiment it would have proved 

 that this insect did not exploit its plant to the full, and yet that what 

 was left would not by itself be enough for another insect, whereas 

 it would suffice if eaten with what was left of other plants. Thus 

 insects might be intercalated which would have to feed upon three 

 or four kinds of plants, and then insects which ate whole groups or 

 families of plants, and finally insects which went everywhere and 

 ate whatever there was to be eaten. 



At the same time, we must remember that no species of plant 

 can be entirely devoured. So much of it must always be left that 

 it can still flower and bring forth its fruit, in order to spring up 

 anew in the following year, otherwise the insects of the next genera- 

 tion would find nothing to eat. And for this reason the relation of 

 the insect to the plant must be a more subtle relation. The number 

 of eggs laid must be exactly proportioned to the amount of nourish- 

 ment provided. Small plants which grow singly in the midst of other 

 plants receive always only a few eggs. The butterfly flies on until 

 it comes to the next plant of this species, and again lays two or three 

 eggs, and so forth. Trees, on the other hand, which contain a surplus 

 of food in their massive crowns, may be favoured with all the eggs 

 of the insect whose larvae are adapted to live on them. The Brazilian 

 butterflies, however, are often very large, and so are their cater- 

 pillars. In order that these caterpillars, when feeding in such 



192 



