in the British Museum. 139 



groove bordered with thin yellow hairs, are both fused with 

 the right half of the fourth tergite, which is rather short, 

 especially laterally. The right half of the fifth tergite is 

 consequently without connexion with its corresponding half; 

 it is short, almost triangular, and bears the usual marginal 

 row of yellow hairs. 



3. Cerceris quinquefasciata. — The segmentation shows 

 dorsally very remarkable variation. The first segment is 

 normal ; the left half of the second tergite, which bears a 

 moderately large yellow spot, merges into the right half of the 

 third, being rather short and almost yellow, on account of 

 its prominent spot. The right portion of the second segment, 

 which is very small and with a tiny yellow spot, is well 

 limited from the tergite thus established. The left half of 

 the third tergite, which is almost yellow, is well separated by 

 a gi'oove from the left part of the fourth, but passes into its 

 right half without any limitation. The fourth segment, 

 together with the one half of the third, becomes thus a split 

 segment {cf. Morgan, 19, p. 245). 



Many cases of a similar kind of abnormality or variation 

 in the segmentation have been described in diiferent forms 

 of Annelids (compare the literature, 18-21). In outer 

 appearance many of these abnormalities (Cori, 18, Taf. xxv. 

 figs. 5 a, b) are very like those described here, especially in 

 Megachile (segments iv. and v.) and Cerceris (segments iii. 

 and iv.) , and also in Chelifer javanus, Thor. (PI. VIII. fig. 1 a) . 

 Cori (i 8, p. 576) characterizes the abnormality in the following 

 manner : — " Dass die ausseren Segmentgrenzen der betref- 

 fenden Korperabschnitte nicht die Form von Kreistouren 

 haben, sondern Schraubenlinien beschreiben." It has also 

 been called "spiral segmentation''^ (Bateson, 21). 



Even if there is great similarity between the more simple 

 cases of *' spiral segmentation " in the Annelids and the 

 abnormalities found in the segmentation of the above- 

 described Articulata, it remains to be proved that the 

 similarity is more than superficial. The number of the 

 segments is not so constant in the Annelids as in most 

 Arthropods, and the individuality of each segment, especially 

 in the inner structure, is more pronounced in the former 

 than in the latter group. The distinction between dorsal, 

 ventral, and lateral parts within each sternite is better marked 

 in the Arthropods. All these are reasons to be careful in 

 comparing and identifying similar phenomena of this kind in 

 the two groups. Different authors have tried to explain the 



