4 Earlc C. Stevenson 



with c. By this method it is determined whether one character 

 a, regarded as "subject," varies in an equal or approximate ratio 

 with its relative b or c. Each hook when measured was num- 

 bered and a corresponding number placed opposite the record of 

 magnitudes of that hook, making it possible to compare the draw- 

 ings of any two hooks whose magnitudes presented any par- 

 ticular deviation in one or more of its characters. The large 

 hooks of each species were considered apart from the small hooks 

 of that species. 



In the seriation or classification of data the varying magni- 

 tudes of any one character were grouped into classes with a class 

 range of 3 microns ; all measurements were made in microns. 

 This minute distinction may seem arbitrary and over-precise at 

 first thought, but when it is noted that 3 microns actual value are 

 represented on the magnified drawing by 1.75 mm. when No. 8 

 objective was used, and 1.33 mm. when No. 7 objective was used, 

 the practice is seen to be perfectly rational. For the greater mag- 

 nitudes of the hooks a wider class range could be used, but in 

 considering the smaller character, d, a smaller class range is im- 

 perative. From these arranged data the mean or abscissa of the 

 center of gravity of the magnitudes of the total number of hooks, 

 and the mean magnitude of the total number of measurements, 

 can be computed. The limits of the magnitude variation of any 

 one character are also established. 



With the method of study thus briefly outlined, the develop- 

 ment of the hooks should be reviewed, since the question of 

 growth of the hooks influences the significance of the results ob- 

 tained in this study. 



POSITION AND ATTACHMENT OF THE HOOKS 



(Figs. 47, 48, 49, PI. VI) 



In the two species of tapeworms under discussion the hooks 

 are restricted to a circular margin on the very short rostellum, 

 within the region bounded by the four cup-shaped suckers that 

 are located one at each corner of the square head. The rostellum 

 in these parasites is very short and rudimentary (see fig. 47, pi. 



194 



