380 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXII. 



consequently not so easil}^ noticed when feeding in wood, ^^'hile the 

 presence of the former can at once be detected. This is rendered 

 all the more easy for the folloAving reason. Nearly all Termites — 

 except the grass-feeders, like .Eutermes, etc. — avoid light and 

 air as much as possible. Now Odontotermes Fece feeding on the sur- 

 face of a board or so, wonld necessarily be exposed to these two 

 from the termite-standpoint undesirable elements ; so to guard 

 against them the White Ants build a cover of mud taken from the 

 surrounding ground over the space chosen for the attack, and then 

 start eating the wood underneath the shelter. These coatings of 

 mud — which are either red or black or greyish according to the 

 colour of the soil — form at times patches of considerable extent over 

 the surface of the wood, and cannot be possibly overlooked by any 

 tolerably attentive observer. 



It should be noted that both Leucotermes and Coptotermes con- 

 struct covered passages as well when for some reason or other they 

 are forced to appear on the surface, v.c/., when trying to get at 

 wood some distance away from their underground nests. But these 

 passages have always the appearance of lines, never of broad sheets 

 as with Odontotermes Fece. They are just large enough to allow 

 the Termites to pass each other when meeting in the tunnel. 

 Moreover, these passages are constructed of the same material as the 

 fillings, that is to say, they consist for the most part of excrements 

 of Termites ; in no case are they made of earthy material taken from 

 the ground near by. The tunnels of Coptotermes are slightly larger 

 and their building luaterial looks somewhat more coarse, than with 

 Leucotermes. 



The feeding figure of Odontotermes Fere is represented in the 

 second and third photographs of plate IV. Its characteristics are 

 that the wood of the whole area on which the Termites have been 

 feeding, is destroyed from the surface iuAvards, not even the hard 

 portions of the annual rings being left intact. It is, moreover, a 

 striking fact that usually not the entire surface is attacked by 

 Odontotermes Fece, but only certain portions of it (cp>. especially the 

 last photo). Thus the wood presents the view of an irregular 

 arrangement of holes — some deep, some shallow — interspersed with 

 narrower or broader patches of untouched wood. 



In the second photo we see parts of three deal-wood staves taken 

 from a cement-barrel. The middle piece is an instance of a rather 

 exceptional case : the entire siu'face of the wood has been attacked 

 and eaten up to various depths. The other two pieces show what 

 commonly happens : in some places the surface area has been left 

 intact (the dark patches appearing in the picture), in others it 

 has been destroyed together with portions of the wood-tissue 

 underneath (light patches of picture). 



On the third and last photo a big branch cut down from a live 



