Edwar Albert Butler, Pond life insects, Londres, Swan Sonnenschein Lowrey, 1886, p. 67-69.

The nature of the external adornment will depend upon the species we have secured, and upon the materials that may happen to have been obtainable by the larva. Small bits of stick, rushes, roots or fibres, blackened by long soaking in the water, and completely water-logged, grains of sand or small stones, the leaves of trees or fragments of aquatic vegetation, the seeds of rushes or other plants growing by the waterside, and the shells of fresh-water mollusca, both dead and living, – these are some of the principal materials employed, their exact nature and arrangement being determined by the species, and sometimes even by the age, of the larva. Some will cut little shreds of vegetable matter all of the same length, and arrange them side-by-side in a spiral manner with wonderful regularity; others will take whole leaves of poplar, willow, and other trees, and attach them flatly to the case. Some will select small stones and stick them on with great dexterity, forming a tube which reminds one forcibly of the exquisite structure made by the marine worm Pectinaria belgica, which occurs not infrequently on our sandy sea-shores. Others will strengthen their tube with very fine grains of sand, making a case in shape like an elephant’s tusk. Some will select straight bits of stick or rush, and place them longitudinally, when they will sometimes project far beyond the ends of the case, like handles; others, using the same materials, but in shorter pieces, will place them transversely, putting each piece tangentially to the surface, so that the ends form a perfect chevaux-de fries round the case, which, if looked at down its length, reminds one of a stocking, carrying set after set of the needles with it has been knitted. But unquestionably the most interesting are those that are adorned with shells. Caddis-worms are excellent conchologists, and by obtaining a number of their cases, you may get together a very respectable collection of fresh-water shells. Sometimes you get the same shell throughout, when the case is often extremely elegant and symmetrical; but frequently you may find five or six species on one case, and then, if the shells are very dissimilar, of course the symmetry of the structure is destroyed. The most elegant are those formed of the smaller species of Planorbis , flat, spirally-coiled shells, something like tiny snakes rolled up. Of these elegant little objects sometimes as many as fifty specimens got to adorn a single caddis case. Then there are the smaller kinds of Limnoea, conical, spirally twisted shells of delicate texture, one or two of which may sometimes be found filling up odd corners, while, projecting here and there, like so many excrescences, may perhaps be seen the stouter and broader shells of Bythinus, the mouth of which is closed by a sort of trap-door. Again, we may find the much smaller and more depressed spiral shells of Valvata, which, with the spires all turned inwards, sometimes compose almost the whole case; and lastly, stuck in here and there wherever there is room, the simply conical abodes of the tiny fresh-water limpets belonging to the genus Ancylus. But besides all these, there are the shells made up of two similar parts hinged together, – bivalves as they are called, – belonging to the genera Sphoerium and Pisidium; sometimes a single valve is used, but more frequently the pair, especially of the very common species called Sphoerium corneum. This is a tolerably bulky shell, and often exceeds in diameter the case which it adorns, so that if three or four of them are used on one case, it acquires a very irregular form. It is not always dead shells that are chosen; very frequently living molluscs are made use of just as they are, thought their consent to the arrangement does not appear to be sought, and the plans of their life must be very greatly interfered with by this unceremonious attachment. Mr. McLachlan, the historian of the European Trichoptera, says that he has seen the wing-cases of water-beetles sometimes mixed with other things as ornaments to the cases, and even the cases of other and smaller caddis-worms, and that, too, while they still contained their inhabitants. The means of attachment of all these objects is the same silky secretion that lines the tubes.
The operations of the insect in the construction of its domicile are very interesting, and may be watched by any one who will take the trouble to eject one from its dwelling and provide it with materials for the formation of another. The two following instances are from the records of the continental observers Meyer and Pictet. The first refers to the formation of a vegetable case. A larva, deprived of its case, seized a piece of reed, and bit off from it a portion of the requisite length; then, cutting a slit in one side, it crawled in and closed up the rent with silk and vegetable débris, and there was the case, fully made. When pieces of reed too short for the case were intentionally given to it, it pieced them out to the required length by cutting off fragments of leaves and attaching them to one end. The other refers to the formation of a mineral case. The larva collected two or three smooth stones of moderate sizes, and made a low arch by fastening them together with silken threads; then, placing itself under this arch , it took up one stone after another , and with its feet fitted them in as carefully as a bricklayer would lay his bricks, attaching them to the neighbouring stones when satisfied as to their position. The stones were always placed smooth side inwards. In this way it took between five and six hours to complete the case.
If the case should be made too long, pieces are cut off till the right length is obtained. As not only the length, but also the width of the case, is always suited to the size of the animal, it becomes interesting to inquire how provision is made for growth in diameter; as the creature grows, each new circle added at the anterior extremity is made o rather larger diameter, thus giving the whole tube a somewhat conical shape; then the smaller end is cut off, and so by repeated additions to one end and subtractions from the other, the case is always the right size, and thus, too, one can understand how it is that a caddis which begins life with a leafy case, may perhaps end it with a stony or shelly one, and that, too, without ever quitting its tenement. Some species do not seem to be at all particular as to the materials they use, but others are so fastidious that they will rather go unclothed (which, of course, means speedy death) than adopt the wrong material.
The cases hitherto referred to are free, and the larva drags its abode about with it, as it crawls slowly along, with just so much of its body projecting from the case as carries the three pairs of legs. But many, especially of the smaller species, and those that live in very rapidly running water, make cases which are attached to stones, and consist of oval, irregular masses of fragments of stones. Some, again, live in company under a common covering of vegetable débris fastened together with silken threads, while others form on the surface of large stones silken canals covered with slime and mud.