132 DIRECTIONS FOR COLLECTING, ETC. 



horizon. I have generally observed a coppery or yellow sunset to foretell rain ; Init as an indication of wet 

 weather approaching, nothing is more certain than a halo round the moon, which is produced by precipitated 

 water: ami the larger the circle, the nearer the clouds, and consequently the more ready to fall." I must not 

 omit, in conclusion, his beautifully simple verification of the rustic couplet, 



" A rainbow in the morning, is the shepherd's warning : 

 A rainbow at night, is the shepherd's delight." 



" A rainbow can only occur when the clouds containing or depositing the rain are opposite the sun ; and in the 

 evening therefore the rainbow is in the east, and in the morning in the west ; and as our heavy rains in this 

 climate arc usually brought by westerly winds, a rainbow in the west (occurring only in the morning) indicates 

 that the bad weather is on the road, bv the wind, to us ; whilst a rainbovr in the east (occurring only in the 

 evening) proves that the rain in these clouds is passing from us." 



These remarks of the philosophic fly-fisher, beside the general information they convey, may teach the young 

 entomologist how to select his weather with a good chance of a fine day ; and also that popular omens are not 

 to be rejected at once by superficial and pert reasoning, but that they are generally founded on truths, however 

 deeply concealed by an accumulation of fancy or superstition. Having stated that a favourable day is indispen- 

 sable to a successful search for insects, more particularly as I am now referring principally to butterflies, as the 

 only class of insects treated of in this volume ; it is next necessary to suggest the best seasons for search. Lonn- 

 then before any specimens are to be taken in the winged state, the collector may, as early as the end of January, 

 dig for the chrysalides of such species as enter the earth to undergo their transformation to the pupa state ; 

 and these he will be most successful in finding near the roots of such plants as the caterpillars feed upon. Other 

 species he will find upon walls or paling, near the food of the larvw ; and others still attached to the withered 

 stems of the plants of the previous summer. lie may also search for the eggs of many species, the most likelv 

 places to find vvhich, and the various modes of depositing them, he will find described in the body of the work. 

 Caterpillars may be collected as early as the beginning of April, and the best time to find them is early in the 

 morning and late in the evening, or even night, as many species remain concealed during the greater part of the 

 day, and some feed only at night ; consequently, a search for them by day would be fruitless ; though some might 

 occasionally be found by pulling up the plants and carefulh' examining the roots, about which they sometimes 

 lie concealed. Caterpillars may be sought all through the summer, and as some butterflies are what is called 

 douijle-brooded, their lar\re are to be found as late as September. It will be useless here to repeat at what parti- 

 cular season each species is found, as that will be found fully described in the preceding pages. I therefore merely 

 remark in addition, that some that can scarcely ever be seen, as they feed at the top of high trees, may be taken 

 by shaking or beating the tree — such, for instance, as the beautiful larva of the purple Emperor, which feeds upon 

 the highest branches of the oak, which is indeed a fertile theatre for the occupation of the entomologist, each 

 tree affording shelter and food to various tribes of insects, too numerous to specify. A white cloth or sheet should 

 be spread upon the ground bi'foro beating or shaking trees. [These observations are as applicable to moths as 

 to butterflies.] 



Wherever the collector is a draughtsman, a careful and exceedingly accurate drawing of several individuals of 

 every species of caterpillar should be taken, and each caterpillar kejit separate, and distinguished by a numler. 



