ON THE VARIATIONS OF AGRION PUELLA. 331 



The variations in the shape of the spot on the second seg- 

 ment, numbered 4, 5, and 6, are represented in the De Selys 

 Collection, at Brussels, and the specimens are labelled "Lch.," 

 no doubt an abbreviation of Longchamps, the name of De Selys' 

 home near Liege. Variations 4 and 5 are also figured by De 

 Selys in Bull. Acad. Brux. x. (2) p. 162, fig. 3 (1843). 



Variation 4. — Sometimes the lateral branches of the spot are 

 entirely separated from the transverse line ; this modification 

 resembles the spot in A. lumdatum, male, and also in an 

 analogous variation of A. hastidatiim. Another specimen of this 

 variety was taken by Mr. H. J. Watts at Lechlade, Gloucester- 

 shire, on June 15th, 1910, and is now, through the courtesy of 

 Mr. Watts, in our own collection. 



Variation 5. — More rarely the transverse line is strongly 

 interrupted in the middle, and the spot then consists of two 

 distinct portions. 



Variation 6. — Again, the transverse line may carry a third 

 branch, lying in the median position between the usual lateral 

 branches ; the spot is then converted into a trident. A male 

 taken by ourselves in Epping Forest on July 14th, 1901 (Entom. 

 1902, p. 37), exhibits a similar variation, as well as other 

 abnormalities, but in that specimen the trident-mark is neither 

 so clear nor so symmetrical as it is in De Selys' beautiful 

 example. 



Variation 7. — The figure of Charpentier's Agrion furcatum, 

 male {=A. puella, Linn.), in * Libellulinse Europsese ' (pi. xl.), 

 shows the U spot connected with the posterior circlet by two 

 widely separated curved lines. It is true that in this respect 

 the figure is in conflict with the description — " [Segmentum] 

 secundum macula furciformi, cum margine postico nusquam 

 cohserente" (p. 158) — and that De Selys characterized it as 

 " inexacte," but we have at least two males agreeing fairly well 

 with the figure. 



Variation 8. — The proportions of blue and black in the 

 coloration of segment 6 vary a good deal, and the segment may 

 be mostly blue or mostly black. 



Variation 9. — The petasus-like black spot on segment 9 

 occasionally undergoes a good deal of change in shape and size ; 

 sometimes, for example, it may nearly cover the entire segment, 

 and at other times it may be represented merely by a pair of 

 separate lateral comma-like marks. 



Cases are not infrequent in which supernumerary black dots 

 and lines, irregular in outline and position, appear upon various 

 parts of the abdomen. Such markings seem to be of a more 

 accidental character than the variations previously considered, 

 and may be so numerous as to cause a more or less general 

 blackening of the insect, as in the cases recorded by us in 

 Entom. 1909, p 294. 



