4:6 



SCIENCE 



[N. S. Vol. XXXIII. No. 847 



in accord with that we had indicated with 

 larvse.* It is now clear that the old classifica- 

 tion, employing the relative length of the 

 palpi in the two sexes as a primary character, 

 is not a natural one. It is true that the 

 length of the palpi indicates, in a general way, 

 the evolutionary state of the species. The 

 forms with the long palpi are the more primi- 

 tive ones, reduction having occurred first in 

 the female, where, on account of the feeding 

 hahits, they were most in the way, and then 

 in the male. The forms with the palpi short- 

 ened in both sexes are therefore evolutionally 

 the highest. 



Unfortunately for any system based on this 

 character, the redviction of the palpi has taken 

 place independently at different points and on 

 this account the forms with the palpi short in 

 both sexes can not be associated, as was for- 

 merly done. Dr. Adolpho Lutz was the first, 

 in 1904, to subordinate the palpi and to out- 

 line a natural classification. Theobald has 

 adopted this classification as the frame-work 

 for his complicated scale-character system. 

 We have found good reason to go still further 

 than Dr. Lutz and discard the palpi for even 

 generic limits. Aside from this it will be 

 found, when the errors and weak points are 

 eliminated from both, that the classification of 

 Dr. Lutz, and the one we formulated on larvae 

 alone, agree very closely. Furthermore, the 

 genitalia, when understood, indicate the same 

 grouping. It seems that the question of sta- 

 bility strongly infl.uenced Dr. Felt in his criti- 

 cism. We venture to point out that only by 

 attacking problems from new' points of view 

 can we progress. There are no fences in true 

 science. 



The further salutary effect of the study of 

 the larval characters has been to establish a 

 more exact and homogeneous concept of gen- 

 eric values. The best example is the genus 

 Culex. Formerly the most heterogeneous ele- 

 ments were united under this name, and with 

 those who work with the superficial characters 

 of the images alone this is still the practise. 

 Thus, even in the last volume of his work, 



* " On the Classification of the Mosquitoes," 

 Canad. Entom., Vol. 39, 1907, pp. 47-50. 



Theobald associated wholly unrelated forms in 

 this genus, while, on the other hand, forms 

 which should be included are removed and 

 scattered through the system on account of 

 trifling difi^erenees in the scale vestiture^ We 

 believe that an impartial and careful study of 

 the Cullcidse from all points of view will show 

 that we were justified in overturning the 

 crude ideas on which their classification was 

 based and that this was brought about through 

 our study of the larvae as organisms unrelated 

 to their adults. We believe that the names 

 founded on larvas are valid and should be 

 given due priority. The characters used are 

 in many cases more positive and reliable than 

 those found in the adults and are further of 

 more evolutionary importance than those here- 

 tofore used in classifying the adults. 



Harrison G. Dyar 

 Frederick Knab 



the germ-cell determinants of chrysomelid 



EGGS 



There lately appeared in these columns 

 under a title similar to the above, an article ^ 

 by Dr, R. W. Hegner dealing with a paper 

 of mine " in which I had questioned his use 

 of the term " germ-cell determinants " as ap- 

 plied to the granules of the pole disc of chrys- 

 omelid eggs. As I do not quite agree with 

 Hegner's interpretation of my position, I take 

 this opportunity to make a more explict state- 

 ment of my point of view. 



To my mind, a cell or tissue determinant 

 suggests or implies a physiological activity 

 that, to use a crude illustration, resembles the 

 physical action of a dye upon impressionable 

 metal. Thus a germ-cell determinant would 

 be something that stamps the undifferentiated 

 cells arising from the cleavage nuclei with a 

 specific germinal or reproductive property. 



I have given evidence that what he calls 

 " germ-cell determinants " are in all likeli- 

 hood merely a part of the food stream from 

 the nurse cells, and this Hegner admits is no 



' R. W. Hegner, " The Germ-eell Determinants 

 in the Eggs of Chrysomelid Beetles," Science, 

 Vol. XXXIII., No. 837. 



' Biol. Bull., Vol. XVIII., No. 4. 



