102 MODERN CLASSIFICATION OF INSECTS. 



it to seize objects above rather than in front of it; so that when 

 an object is perceived floating on the surface of the water, the larva 

 rises very cautiously until it has nearly reached it, when, by a sudden 

 jerk of the neck, it seizes the object with its jaws, and immediately 

 drags it under water; if it still struggle, the larva endeavours to dispatch 

 it by repeated jerkings of the head. When in the water they may con- 

 stantly be perceived jerking themselves in every direction, probably 

 for the purpose of seizing upon other minute insects. 



The larvae of the genus Colymbetes differ from the preceding in the 

 large size of the quadrate anterior segment of the body ; the two 

 caudal processes are also much longer. But the most singular larva; 

 in the family are some of small size (^fig- 6. 5.), which I have found in 

 the neighbourhood of London, and from which the Rev. F. W. Hope, 

 (who has paid great attention to the tribe of Water-beetles) informs me 

 that he has reared Noterus crassicornis. They are of an elongate oval 

 shape, attenuated towards the posterior extremity ; the head {^jig. 6. 6. 

 seen from above, 6. 7. seen sideways) is produced in front into a 

 slender beak, longer than the rest of the head, the jaws meeting to- 

 gether at its tip ; the terminal segment is furnished with the two long 

 ordinary filaments, between which the segment itself is produced into 

 along and acute point. The maxillae and labium are represented with 

 their appendages in Jig. 6. 8. These larvae are brown, with two 

 ochreous bands across the back. They are about a quarter of an inch 

 in length. I have also found another larva of a smaller size than the 

 preceding, in which the produced part of the head is shorter and 

 more robust, the terminal segment produced but very slightly between 

 the caudal filaments, which are very long : this larva is of a dull pale 

 colour, with two irregular, longitudinal, dark markings down the back. 

 As to the affinities of the present family with the Carabidse on the 

 one hand, and the Gyrinidoc on the other, entomologists seem to be 

 entirely agreed. Many links in the chain, it is true, remain to be dis- 

 covered, the larvae being also of a totally different appearance. 



Mr. MacLeay considers that an obvious analogy exists between the 

 larvae of this family and of the Neuropterous genus Hemerobius, and 

 as the former are not furnished with the lateral branchial appendages of 

 the Gyrinideous larva?, their appearance is much less scolopendriform 

 Clairville, Leach, Erichson, and others, have made this family the sub- 

 ject of especial memoirs ; and by these authors various plans of classifi- 

 cation have been proposed, and upon which the distribution of the British 



