the chief reflection must have been from the stem, or subterranean mass,from which the superficial portion was derived.

Passing on now to the consideration of the second type of compactmasses of trachyte 4 , we are met by the difficulty of not knowing their exactform and position; there is little doubt that they generally assume the formof vertical plates, since, where exposed by denudation or otherwise, thisform is clearly indicated. In volcanic cones we know that these masses fillfissures which tire usually radial to the central axis of the main cone.Nevertheless at the Monte Impcralore some horizontal dykes exist. Such massestis these may produce most remarkable effects in modifying both theemergence and azimuth of an earth-wave , as is well illustrated in the N.AY. corner of the island. Also, the fact of the activity of the shock di-minishing much more rapidly in an easterly than in the opposite direction,may be due to the presence of three or more such walls of rock as havesupplied the craters of Porto d’ Ischia (?), Montagnone, and Rotaro.

With regard to the latter fact , the less elastic nature, the intricateinterweaving of rocks of different elasticity, resulted in continual reflectionand refraction, in addition to the, absorption by the loose materials.

The value of a completely incoherent rock, as a material for foundations,was remarkably illustrated at the marinas of Lacco and Casamicciola , forthe houses built on the beach, (the foundations of which were often lickedby the waves), were so slightly fractured that the damage done to them wasout of all proportion to their distance from the seismic focus. It is said thatall rules have an exception, and certainly that one indicated in the well knownparable relating to the building of a house on a rock has one in the caseof earthquakes.

Physical configuration had much to do also with modifying the - effectsof the earthquake. The island of Ischia , from its mountainous character,possesses few level spots, and as a consequence , many houses were builtnear the edge of a scarp, which was often of loose tula , and ill supportedby revetment walls or oilier means. As a consequence of this , partial orcomplete landslips were very abundant, so that many houses and buildingswere destroyed from this cause , their foundations actually slipping frombeneath them. Numerous fissures occurred in the injured district; but thesewere always parallel to some scarp edge , and were the result of partiallandslip.

There were very few buildings fluff failed to show more thanthe effects of one shock. Besides the geological structure, the physicalconformation seems to have done much to reflect the wave-paths continually,so as to produce that rotation of objects of which two remarkable instancesarc those illustrated in Photos. IT, XVIII, and Figure 2, Plate I.

My own opinion is, that mountains and lulls are subordinate to rockstructure and density in their power to reflect the vibrations, although the