122 PROCEEDINGS OF THE NATIONAL MUSEUM. 



which the insect moves very little or not at all, and takes no food, 

 'llic wing-s develop entirely during this stage and are outside the body 

 skin. The metamorphosis approaches closely to a complete one, but 

 on account of the similarity of larval and adult forms and mouth parts 

 it nuist still be considered as incomplete. Reproduction is oviparous 

 and fi-equently parthenogenetic. 



MP^.TFIOD OF MEASUREMENTS. 



A few of the descriptions of Thysanoptera previously published 

 have been found to be too brief and general for the recognition of the 

 species. Another difficulty which has been noted in some descrip- 

 tions is the giving of comparative dimensions relative to other spe- 

 cies. This may be useful to the collector if he happens to have or 

 know all the species referred to; otherwise he is at an utter loss to 

 know what is meant. Having exjierienced these difficulties at various 

 times, the writer came to the conclusion that each description slioukl 

 be al)solutely complete in itsf^lf and independent of all others, and that 

 therefore a system of measurements based entirely upon the species 

 under consideration would frequently prove of service in the determi- 

 nation and separation of these insects. The eye can not be relied 

 upon for exactness in this matter, as has been frequently found in the 

 course of this work, and therefore all measurements given in tlic fol- 

 lowing descriptions have been made in the same way, by means of an 

 eyepiece micrometer, as follows: A stage micrometer of reliable make 

 was first pro^-en to l)e accurate b}^ comparison with a steel millimeter, 

 scale, then with each combination of lenses used the num])er of spaces 

 on the image of the stage micrometer covered b}^ the scale of the e}'e- 

 picce micrometer w^ere determined, two points being selected where 

 the divisions coincided. Then the number of spaces covered on the 

 stage micrometer w^as divided by the number of spaces of the eye- 

 piece micrometer covering tliem, and the quotient was, evident!}', the 

 fraction of a micromillimeter upon the stage shown by one division of 

 the eyepiece micrometer. This quotient may be called the factor of 

 the eyepiece micrometer for that combination of lenses and will hold 

 unchanged for any object measured w^ith that magnification, ))ut wdll 

 of course vary for every other magnification. Illustration: Using a 

 1-inch objective and a 1-inch eyepiece (Bausch and Lomb) with the 

 tul)e closed, I find that the fifty divisions of the eyepiece micrometer 

 cover, say, exactly 1 millimeter of the stage micrometer. Dividing 

 then 1 millimeter by 50, I have two one-hundredths millimeter, which 

 is the factor for that combination of lenses. Now, placing the object 

 to be measured upon the stage, we find, e. g., that forty -five spaces of 

 our eyepiece micrometer just cover the object to be measured. INIul- 

 tiplying by the determined factor, we have two one-hundredths milli- 

 meter times forty-five, which equals ninety one-hundredths millimeter 

 as the length of the object measured. This method has been used in 



