Notes in 1887 upon lepidopterous larve, éc. 577 
remarkable, as the species are so closely allied. In 
Diagram II. (p. 578) the rates of growth in these two larve 
are shown by the graphic method. It must be remembered 
that the mature S. convolvuli certainly attains a much 
greater length than that indicated by the ordinate VI’. 
Thus Mr. Buckler describes a full-grown larva as 4 in. 
~ in length,—more than one-fourth as long again as the 
average length of my mature larve. As, however, my 
measurements are the only ones available for the other 
stages, it is better to retain the length given in the 
diagram. The curve thus represents the true lengths 
measured during the life-histories of one set of larve. 
I believe, however, that the lengths of the other stages 
are about normal, and that the effect of artificial sur- 
roundings, or the naturally small size of my larvae, 
became manifest in the last stage only. The lengths of 
the larvee and the duration of the periods are those given 
at the end of the account of the ontogeny. The corre- 
sponding data for S. ligustri are taken from my paper in 
Trans. Ent. Soc. Lond., 1885, p. 281 et seq. The length 
of the last stage (75°0 mm.) is taken from Mr. Hellins’s 
account of the larva in Mr. Buckler’s volume (J. ¢., p. 111). 
The length of the larva at the beginning of the 5th stage 
is estimated at 36°0 mm., as the measurement (33°0 mm.) 
given in my paper was taken from the larva in the Sphinx 
attitude (l. c., p. 288). The data for Aglia tau are to be 
found in the present paper. 
In order to facilitate comparison between the growths 
of the two Sphinx larve, the lengths of S. ligustri have 
been marked on the ordinates of S. convolvuli, and con- 
nected by a dotted line, and conversely those of the 
latter upon the ordinates of the former. If the duration 
of the larval stages were the same, then the comparison 
between the dotted and continuous lines would represent 
the real relation between the larval growths. I have 
already given reasons for the belief that the duration of 
larval life in S. convolvuli is abnormally long in this 
country, and it is therefore possible that the dotted line 
which crosses the ordinates of S. ligustri more nearly 
represents the normal condition of S. convolvuli than the 
line which represents the observed condition of the latter 
species. 
The most remarkable fact shown in the diagram is the 
manner in which the rate of growth in S. convolvuli 
