24:6 THE ENTOSIOLOGIST's RECORD. 



galls ; luy own impression is, that 24 hours is nearer the truth in a 

 season like this, when the budding of the oak is delayed till late, and 

 then takes place in weather more like July than early May. Up to a 

 certain point the gall-bearing 1 )ud expands like a normal healthy one : 

 then it assumes a rather rounder form, biit, as a similar form is largely 

 affected by buds containing catkins, this is not a very helpful indi- 

 cation. In the next stage, the top of the gall is just visible among the 

 scales at the apex of the bud, and 1 could probably have detected these 

 had I been as learned when I began the search, as I was when I ended 

 it. Still, they differ very little from many catkin buds ; the chief differ- 

 ence being, that the green surface of the exposed gall is smooth, whilst 

 the catkin where exposed is nodulated and of a yellow-green. In a very 

 few hours the gall is more exposed, and is then toleraljly easih' spotted 

 after a little jDractice ; it is fairly firm and solid, and one has little fear 

 of injuring it when breaking it off Avith the bud containing it. In a very 

 short time it expands to the size of a small pea, and the fly escapes. 

 The gall still further enlarges after the escape of the fly, but now 

 readily gives way under the finger in breaking it off. The 

 cavity within is very large — almost big enough for the fly to buzz 

 freely in it — but the enlargement is so recent, that, up till the last few 

 days, the Neuroterus must have been as closely confined by his lodgings 

 as any other gall tenant. The flies that so emerge are J and 2 , so 

 that this stage of the insect corresponds to the summer gall stage of 

 those species of the genus that are more strictly entitled to be called 

 spangle gall makers. 



There is no difficulty in collecting a reasonable supply if one 

 happens to hit uj)on the right tree at the right time, but to ensure doing 

 this, much time must be spent in looking daily over a nvimber of trees 

 and watching their progress. I was very lucky, therefore, to secure even 

 such a measure of success as I did, not liaving given much time to the 

 chase, and starting in great ignorance of how the hunt should l^e con- 

 ducted. In a couple of days more the gall shrivels up, and the bud 

 returns very much to its normal appearance but with a ragged dis- 

 hevelled aspect. This at least is the case with a very dry atmosphere 

 and hot sun, and with the thermometer at 70°, the oak leaves being 

 nearly an inch long, and the galls of Andricus quadriUneatus and Neuro- 

 terus baccarum quite obvious. What may occur in a cold damp season 

 I do not know ; probably the gall remains fully expanded till it withers, 

 rots, and falls off. 



I have said that the shrivelling of the gall restores the bud to a 

 nearly normal aspect, a result one would not expect on examining the 

 gall when at its greatest expansion ; a comparison of several galls in 

 both conditions, however, explains how this occurs. The gall is really 

 formed in, and consists of, the little cone of succulent stem that forms 

 the base of the bud, and carries on its surface the inner series of brown 

 scales forming the bud ; when at its largest, the green succulent wall 

 of the gall is more freely developed between the bases of the scales at 

 some places than at others, so that one thinks, on casual inspection, that 

 the gall is placed on the growing point of the bud, and has forced the 

 scales aside, and occupies their centre ; but this is not so ; the gall is 

 under the scales, and the small central chaffy scales may usually be 

 found on some part of the gall, tolerably close together, and other 

 scales on different parts of its surface, and these, though now very 



