October 1892.] 



PSYCHE. 



337 



THE NUMBER OF LARVAL STAGES IN THE GENUS NADATA. 



BY HARRISON G. DYAR, BOSTON, MASS. 



In Psyche, recently, I expressed 

 the opinion that species of Nadata had 

 more than six larval stages, which was 

 founded on certain measurements made 

 from Nadata gibbosa. I have not since 

 obtained this species in the early stages,* 

 but have bred another from the egg, 

 which is N. oregoncnsis Butl. In this 

 species, the number of stages appears 

 to be normally six ; but two individuals, 



* Miss Soule has recently written the life history of 

 N. gibbosa (Psvche, v. 6, 197) and found five stages, 

 as did also Dr. Riley (See 5th rept. U. S. ent. com., 

 1S90). Unfortunately, Miss Soule has given no meas- 

 urements of the head, but she has given the length of 

 the larva in all its stages and the numbers she gives 

 correspond well with a series derived with the ratio .60. 

 This does not corroborate my observations (on N. ore. 

 gonensis), as, to do so, a stage should appear lacking 

 between stages I and II, provided that the measure- 

 ments were made at the first of each stage. Miss Soule's 

 figures are 3.16 (=.19), k (=- 2 $)<k (—5°), 3 (=-75)> ^ nd 

 14 (=1.25) inches while the calculated series would be 

 .16, .27, .45, .75, 1.25 in., thus showing no gap in the 

 series anywhere. To suit my observations the newly 

 hatched larva should have measured .10 inch instead of 

 .16 inch. Miss Soule says "not quite 3-16 in.," which 

 is certainly nearer .16 than .10, as the latter would be 

 not quite g in. 



But I do not think the length of the larva is a reliable 

 measurement to take as it is subject to great change 

 throughout the stage, and, even if taken as nearly as 

 possible at the same time in each stage, is subject to 

 inaccuracies through the expansion or contraction of 

 the larval segments. Moreover it takes no account of 

 the growth during either the first or the last stage, ac- 

 cording as the measurements are made at the end or 

 beginning of each stage, and I should hesitate to assume 

 that the growth was always strictly proportional. In 

 fact, I believe that in Nadata it is not so, for double 

 growth seems to take place in the first stage. 



The measurement of the width of head is open to 

 none of the above objections and possesses besides sev. 

 eral advantages not shown by measurements of the 

 length of the larva. 



carefully bred in confinement and two 

 bred in the open air had but five stages^ 

 while another specimen, less carefully 

 reared in confinement beside another 

 bred in the open air on its growing food 

 plant under a net exhibited six stages, 

 but not the normal ones. All the larvae 

 appeared to omit the normal second 

 stage, even those that had six stages. 

 These latter inserted an extra stage be- 

 tween the fourth and fifth, not differing 

 in markings from the fourth, as will be 

 seen in what follows. The growth 

 during the first stage was very great, out 

 of all proportion to the subsequent 

 growth, and, previous to moulting, the 

 new head in process of formation behind 

 the old one, caused an enormous pro- 

 jection of the body. 



The calculated normal series for the 

 widths of head stands as follows : — I .79, 

 [II 1. 13,] III 1.62, IV 2.31, V 3.3, VI 

 4.7 mm. ; ratio .70. 



From the larvae that had five stages, 

 I obtained the following measure- 

 ments:— 1st, .75, 2nd, 1.55,31x1,2.35, 



4 th ." 3-3, 5 th ' 4-7 mm - 



From those that had six stages : first 



example: — 1st, ,* 2nd, 1.4, 3rd, 



2.2, 4th, 2.7, 5th, 3.2, 6th, mm. :f 



second example : — 1st, ,* 2nd, 1.5, 



3rd, 2.3, 4th, 2.S, 5th, 3.7, 6th, 4.8 



mm. 



* Measurement not recorded. 



t This larva died before moulting the last time. 



