NOTES ORIGINAL AND SELECTED. 93 



to I inch in length. Mr. Cole has since identified the cater- 

 pillars as those of the "Winter Morth " {Cheimatohia hvumata) 

 and the ants as Formica vnfa — the Wood-ant. The ants were 

 very pugnacious, and those with burdens absolutely refused 

 to give up their prize, even when their lives were in 

 danger. The " looper " caterpillars were obtained from 

 an oak tree about seven yards east of the path, and 

 the column of ants was clearly defined, and much 

 narrower in the undergrowth. The ants were scahng the tree 

 in thousands, and had evidently a preference for one particular 

 oak, though two others, a few yards distant, were attacked, but 

 not in such great force. On the other side of the path the 

 column stretched out to an ant-hill, and the total distance from 

 the nest to the oak tree was 15 yards. An ant seized a cater- 

 pillar by a loose fold of skin under the body, and putting its legs 

 astride it dragged its burden along in the same manner as a boy 

 "rides" a toy horse. In cases wdiere the caterpillar w^as very 

 large, two ants set to work, one at each end of the " looper's '' 

 body. The scene of the " raid " was visited on the three 

 following days, and the tiny robbers were found hard at work on 

 each occasion. — H. Whitehead, Essex Museum of Natural 

 History. 



BOTANY. 

 Variation in Adoxa. — The common form of the inflores- 

 cence of the Moscatel [Adoxa moschatellina) consists of five 

 flowers forming a globose head. The peculiarity about the 

 plant is that the flowers have two forms of symmetry, viz., the 

 tetramerous and the pentamerous types, both of which occur on 

 the same plant. In the normal inflorescence the terminal flower 

 is tetramerous and the four lateral flowers are pentamerous. 



Some material was collected during the Spring of 1902, with a 

 view to studying the variation of this plant. Inflorescences were 

 collected at Chiselhurst, Kent ; Caterham, Surrey ; and at 

 Theydon Garnon, Essex ; and it was found that considerable 

 variation occurs both in the number of flowers and also in the 

 number of the parts of the perianth. Out of 1,071 inflorescences, 

 only 934, or 87*2 per cent, possessed the normal number of 

 flowers, viz., five. The number of flowers per inflorescence 

 ranged from three to ten. 



The number of divisions of the corolla varied from three to 



