(I. 11. FuiiL) 101 



Medium stage (2) 10 — 15 mm, long by 1-0 — 1-25 mm, broad. 



Large stage (3) 17 — 23 mm. long by 1-5 — 2-0 mm. broad. 



It should be stated that no living specimen was ever found to be 

 more than 21 mm. long, and that 19 mm, is about the average length, 

 taking a number of specimens into consideration. The 23 mm. specimen 

 was an extended spirit example. After about a year, the small stage 

 was no longer present, and only medium and large stages could be 

 found in the plant pots. According to most authorities the larval 

 period is supposed to last five years. From observations on the length 

 and rate of growth, I am inclined to place the period at four years 

 rather than five. The small stages, taken in one year from July to 

 October, varied much in size. It was found, after a couple of months 

 (November — December), that a number of these apparently small 

 specimens were really of medium size. It appears from this that the 

 breadth is a safer criterion than the length (which is difficult to ascertain 

 on the living larva). Judging from the size of specimens taken from 

 July to October, and also from the fact that the adult beetles are 

 common from April to June, it may be assumed that the newly hatched 

 larva commences its free existence in July. Again, there are three 

 larval moults, performed in the usual manner, by splitting the chitinous 

 skin along the thorax, and leaving the old exo-skeleton behind. From 

 these facts, and from the rate of growth, it may be assumed that the 

 larva is full grown in three years. Full size larvae are found actively 

 feeding up to October ; but these same larvae apparently do not feed 

 again (at any rate on vegetation), and pupate the next summer. It is 

 thus probable that there are three stages limited by the three moults, 

 and occupying three years. Then there is a period of active feeding, 

 followed by a quiescent condition terminating in pupation, all this 

 taking place in a year. It was found that of the large larvae (17 to 

 23 mm.) a number fed through the winter, while some were not feeding 

 after October, and were not lethargic ; these latter were broader than 

 were the former. This would appear to support the above theory. 

 If the life of the larva is one of five years, then it would seem that the 

 early (small) larval stages would continue over a longer period; this 

 is not confirmed by observations. The rate of growth was found to be 

 of such a uniform character as to suggest that the curve of gi'owth is 

 fairly continuous rather than irregular. It seems to the author that 

 the period of growth is not five years, a period given since Bierkander, 

 b.ut probably requires four years from the egg to pupation. The earlv 

 (small) stage larvae were never found feeding. It is stated that they 



Ann. Biol iii 7 



