The greatest care is necessary in the breeding of Tenthredinidae, as most of them transform under ground and are single brooded, the larvae remaining in the ground from midsummer until the following spring. Nothing but constant care in maintaining uniform moisture and temperature of the soil will insure the success of such breeding. Some species bore into rotten wood or the stems of plants to undergo their transformations, as for instance the Dogwood Saw-fly (_Harpiphorus varianus_). This species, unless supplied with soft or rotten wood in which to bore, will wander ceaselessly round the cage, and in most cases eventually perish. Where a small room can be devoted to the purpose, an excellent wholesale method of obtaining wood-boring insects (_Coleoptera_, _Lepidoptera_, etc.) is to collect large quantities of dead or dying wood of all sorts or any that indicates the presence of the early states of insects, and store it in such apartment. The following spring and summer the escaping insects will be attracted to the windows and may be easily secured. The objection to this method is that, in many cases, it will be impossible to determine the food habit of the insect secured, owing to the variety of material brought together. _The Root Cage._--For the study of insects which affect the roots of plants a root cage has been devised by Prof. J. H. Comstock which is of sufficient importance to warrant full description. It consists of a zinc frame (Fig. 125_a_) holding two plates of glass in a vertical position and only a short distance apart, the space between the plates being filled with soil in which seeds are planted or small plants set. Outside of each glass is a piece of zinc or sheet iron (_b_) which slips into grooves and which can be easily removed. When these zincs are in place the soil is kept dark. The idea of the cages is, that the space between the glasses being very narrow, a large part of the roots will ramify close to the surface of the glass, so that by removing the zinc slides the roots may be easily seen, and any root-inhabiting insects which it maybe desirable to breed may thus be studied in their natural conditions without disturbing them. Prof. Comstock has used this cage very successfully in studying the habits of wire-worms, and its availability for many of the underground insects, such as the Cicadas, root-lice, larvae, etc., is apparent. These frames may be made of various sizes, to accommodate particular insects. It will be of advantage in many cases, in order to secure the natural conditions as nearly as possible, to sink the cage in the soil, and for this purpose Prof. Comstock has had constructed a pit lined with brick for the reception of his cages, and employs a small portable crane to lift them out of the ground when it is desirable to examine them.