778 THE AMERICAN NATURALIST. [Vol. XXXVIII. 



head of Blatta gervianica. This was proposed by Prof. Com- 

 stock and to him I am indebted for many suggestions. 



In the choice of an object for this study I was guided by the 

 desire to, select one of the more generaUzed forms and one which 

 at the same time would afford an abundance of material. Both 

 of these conditions seemed to be met by Blatta germanica. 

 Though undoubtedly highly specialized in some respects, the 

 cockroach is regarded by many students as the most primitive 

 of ourpterygote insects. I believe that in this respect one of 

 the Plecoptera would have been more valuable, but as regards 

 abundance of available material it would be difficult to select a 

 - form more satisfactory than is Blatta. 



On the other hand, as objects of study the embryo of Blatta 

 present certain difficulties which are accentuated in the course 

 of the work. Of these there are " die ausserordenthchen tech- 

 nischen Schwierigkeiten " which Cholodkowsky has so empha- 

 sized. These are due primarily to the great mass of yolk, which 

 becomes so brittle as to make it almost impossible to secure 

 complete series of sections by ordinary methods. A more 

 serious difficulty in the work under consideration was the small 

 size of the cells and the indistinctness of the invaginations of 

 the body wall. Of the latter, Wheeler ('89, p. 349) says : " In 

 Blatta the formation of the nervous system in its earlier stages 

 cannot be clearly seen from the exterior. The same holds true 

 of the small tracheal invaginations, though several pairs, especially 

 those of the thorax and basal abdominal rings, may be seen on 

 the plurae in good preparations before revolution. Still they are 

 so much less distinct than in Doryphora that I have given them 

 little attention " 



The question of the number of segments in the body of the 

 perfect msect early attracted the attention of entomotomists. 

 As the history of the development of our knowledge of the 

 number entering into the structure of the head has been very 

 tuUy discussed by Comstock and Kochi (:02), I have but little to 

 add to their account. 



