604 On the artificial Arrangement 



Moreover, when digging vigorously the creature makes a short 

 step backwards, by which means a portion of the sand is 

 thrown on the head: owing to the hump-like form of the back 

 this is immediately jerked away, the body at the same time 

 advancing another step in its backward and spiral motion. 



I placed one of my larva? in spirits ; the other underwent 

 a three weeks' fast during the following months of August and 

 September ; but on my return from the country it devoured 

 flies with its ordinary activity. About the middle of Septem- 

 ber the specimen which had enclosed itself in a cocoon at the 

 end of July, produced a winged ant-lion fly, the skin of the 

 pupa being attached to the edges of a small hole made in the 

 cocoon. The larva which I kept in my study continued ac- 

 tive till the cold weather set in, when it became torpid although 

 near the fire-place ; but at the middle of March, and before 

 the flies were abroad, a few Trichocera only being occasion- 

 ally seen, the ant-lion resumed its habits and made a small 

 burrow ; it died, however, in the following month. 



Art. VI. — On the artificial arrangement of some of the more exten- 

 sive Natural Orders of British Plants. By Frederic John 

 Bird, Esq.* 



Having frequently experienced considerable difficulty in de- 

 termining certain genera and species of plants, belonging to the 

 more extensive natural orders, especially during botanical ex- 

 cursions, when we are necessarily deprived of microscopic aid, 

 I have been led to construct the following tabular analyses of 

 genera, in which many of the more obscure features are disre- 

 garded, and the name of the genus arrived at without having 

 recourse to minute dissection, or indefinite characters. These 

 analyses, drawn up in accordance with suggestions originally 

 made by my brother, Dr. Golding Bird, are entirely restricted 

 to an examination of the floral envelopes and fruit, without 

 any minute dissection of the latter, a process which is often 

 impracticable whilst studying in the field. Two or rarely 

 three characters are taken in each binary division ; such as 

 are most obvious have in every case been selected, and those 

 contained in the first line are exactly the converse of the fea- 

 tures met with in the second ; thus avoiding any serious error 

 which might arise from the too close resemblance in structure. 



* Communicated by Golding Bird, M.D., F.L.S., G.S., &c, Lecturer 

 on Natural Philosophy at Guy's Hospital, &c. 



