40 THE entomologist's record. 



absence of certain tubercles, namely iii and v of mesothorax ailcl vi of 

 alxlomen, according to my numbering, perhaps not the most natural 

 numbering, as Dr, 0. Hofmann pointed out. I have called these sub- 

 primary tubercles. (2) The presence of only single hairs on tubercles 

 which in later stages bear several hairs in the particular species. The 

 violation of either of these points shows the disappearance of the 

 primitive first stage. 



I would remark briefly concerning Dr. 0. Hofmann's criticism of 

 my numbering of the larval tubercles, that, in proposing the numbers, I 

 did not consider the tubercles of thorax and abdomen as homologous, 

 and consequently numbered them differently. Using the mature larva 

 as a basis I numbered in sequence from the dorsal line towards the 

 feet. The first four thoracic sette, being often closely grouped in pairs, 

 I thought to be doubled sette and gave them the same numbers, 

 distinguishing them by letters. Except for the presence of the 

 thoracic shield, which offers difficulties by carrying enough setae to 

 indicate an additional somite, I now agree with Dr. Hofmann that the 

 thoracic and abdominal setae are homologous as follows : 



i a of thorax = i of abdomen and should be called i. 



i b of thorax = ii of abdomen and should be called ii. 



ii a of thorax = iii of abdomen and should be called iii. 



ii b of thorax = iv of abdomen and should be called iv. 



iii of thorax is subi^rimary, not represented on abdomen; it might be called v a. 



iv of thorax = v of abdomen and should be called v. 



Y of thorax = vi of abdomen, being subprimary in both cases, should not bear a 

 separate number, but might be called v b. 



vi of thorax = vii of abdomen and should be called vi ; it is frequently multiple, 

 universally so on the abdomen. 



However, I shall not make these changes in future descriptions, as 

 uniformity in nomenclature is more desirable than a strict indication 

 of homology. Any change would create confusion, which must be 

 avoided. The anal plate does not offer a difficulty like that of the 

 cervical shield. It bears enough setae for a somite distinct from the 

 9th abdominal, but this is clearly shown to be the fact on other 

 grounds, the anal plate representing the 10th abdominal segment. 



It has been suggested to me by several that it would be a help to 

 students to give a model for the description of larvfe. I do not fully 

 agree with this view for the reason that in following such a scheme 

 one would be likely to think that the i^oints there indicated were all 

 that it Avas necessary to observe and thus individual initiative Avould be 

 discouraged in the discovery of new, and perhaps more valuable 

 characters than any yet recorded. However I venture to offer the 

 following as showing the points usually covered in descriptions at the 

 end of the nineteenth century, with the implication that a larva should 

 not be considered fully described if these have not been noted. 



The head : Its shape, whether higher than wide or the reverse, 

 proportions of the lobes, relations of clypeus and paraclypeal j>ieces, 

 comparative length of antenna?, any abnormality in number of ocelli, 

 development of seta? or peculiar processes, presence of secondary hairs ; 

 coloration, describing spots or bands ; position of head at rest, whether 

 erect or fiat, free or retractile in prothorax. 



The hodji : Its shape, Avhether cylindrical or flattened or otherwise 

 modified, noting any humps or peculiar prominences ; segments short 

 or long drawn out and slender ; incisions distinct or otherwise or the 



