796 THE AMEi 



'URAL! ST. [Vol. XXXVIII. 



tnaiKlibular and one in the second maxillary segments. Heymons 

 ('95) basing his conclusions mainly on a study of Forficula, 

 described essentially the same conditions, as applying to Forfi- 

 cula and to the Orthoptera, including the BlattidcX. Carriere and 

 Burger made the same observations on Chalicodoma, thus lending 

 strength to Heymon's surmise that such a mode of origin is 

 typical for the entire group of insects. 

 ^ Various investigators, among whom are Palmen ('77), Hatschek 

 ('77), Wheeler ('89), and especially Carriere ('90), and Burger 

 ('97), have regarded the tentorial invaginations as homodynamous 

 with the tracheal invaginations. To this Korschelt and Heider 

 ('93) object that their rudiments do not by any means everywhere 

 agree so closely with tracheal stigmata of the following segments 

 as they do in Chalicodoma. This is especially true of these 

 rudiments in Blatta. As we have seen, the invaginations for 

 the anterior arms arise close to the median line, just laterad of 

 the nerve cord. This corresponds more nearly to the position 

 of the invagmations for the thoracic furca. The origin of the 

 pos^tenor nivaginations is more suggestive of the stigmatic 



In considering this question it must be remembered that 

 Chalicodoma is a highly specialized type with an interpolated 

 larx-al stage, while Blatta is a representative of a comparatively 

 generalized group of hemimetabolous insects. This being the 

 case, It does not seem that we should regard the location of the 

 tentorial rudiments in Chalicodoma as primitive, but rather I 

 should regard it as secondary. The fact urged by Palman, that 

 the chitmous lining of the tentorium, like that of the trachea?, is 

 shed during ecdysis, is of little weight, since the same holds 

 true of any hollow ectodermal invagination in insects. 



However, it is not necessary to conclude that the tentorial 

 invaginations are metamorphosed tracheae " which have lost their 

 primitive function and become secondarily modified." If, with 



ennel, we derue the tracheae from dermal glands of annelidan 

 ancestors, there is no reason why we should not consider the 

 ^lac ica^ and the tentorium as homologous structures. Certain of 

 trich ^Icfinitely localized have become modified to form 



i^a , while certain others, of the same origin though not 



