OF CHLOËON (EPHEMERA) DIMIDIATUM. 479 
ance of probability, that they would be inconvenient for swimming. However this may 
be, the fact remains, and I was interested to see in what manner the change was effected. 
This is shown at a glance by Plate LVIII. fig. 2, in which we see that the antenna of the 
proimago is not only much shorter, but also thinner than that of the present stage, in 
which it lies like a sword in its sheath. The two basal segments (Plate LVIII. fig. 2), 
indeed, are almost filled by the new organ, but this is far from being the case with the 
thirteen following. In this portion of the antenna (Plate LVIII. fig. 22), the chitino- 
genous layer and the soft internal tissues have contracted very much, so that the new 
antenna lies quite loosely in the old one. It only reaches, however, as far as the fifteenth 
segment. From this part onwards (Plate LVIII. fig. 2 c) the softer tissues of the old 
antenna are unaltered, they take no part in the formation of the new organ, and when 
the skin is shed they are cast off with the dead outer layer of the chitine. The con- 
nexion between the antenna of the proimago and the tissue of the terminal part of the 
antenna is shown in Plate LVIII. fig. 9. 
The mouth-parts do not essentially differ from the description of them given by 
Westwood (* On the Modern Classification of Insects,’ vol. ii. p. 24). 
The upper lip (Plate LVIII. fig. 4) is almost quadrangular, rounded off at the free 
angles, and notched in the middle of the anterior border. It has a number of short 
hairs along the margin, and a few others scattered over the surface. 
The mandibles (Plate LVIII. fig. 5) are short, strong, and horny. The free edge may be 
divided into three very distinct parts. The upper portion is divided into three processes, 
each of which has two or three teeth. The middle portion is occupied by a tuft of hairs. 
The basal part under a low power has a molar appearance, but when highly magnified is 
wen to be set with the curious processes two of which are figured (Plate LVIIL fig. 6) 
one seen sideways, the other looked at from in front. As usual, the mandibles are 
slightly unsymmetrical; and one of them has a strong process at the upper end of the 
molar portion. 
The maville (Plate LVIII. fig. 7) are short, slightly curved, with two strong teeth, 
and à row of spines along the upper inner edge. The palpi are rather longer than 
the maxillæ themselves, and have three joints. According to Westwood the correspo mor 
mg organ of F. vulgata consists of four segments: comparing his figure with my Br 
mens, it seems that the small basal segment is undifferentiated. The two apical segments 
are also less distinct] separated than in his figure. 1 : 
The lower lip (Plate Lv. fig. 8) is membranous and deeply quadrifid. The ^as 
ate three-jointed: but the division between the two terminal segments 13 nel iim d 
marked. Mr. Westwood says that the lower lip is “ furnished within iis J ch addi- 
tongue, of which the anterior angles are produced and pilose ;’ I found, in = - E 
"e the organs already described a three-lobed membranous lip (Plate pin + 44 
Which I presume to be the tongue. It seemed to me, however, to lie ın ibly it may 
mouth and to be connected, not with the labium, but with the labras FT") en 
homologous with the membranous lobes in the Thysanura, which I have doubtfully 
Uggested to represent a second pair of maxillæ (?). he males are slightly 
° small projections on the penultimate abdominal segment of the 902 TRA 
