346 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. C2. 



T. ordinoides^ he states: "I believe that it is impossible to fix the 

 exact eastern and southern boundary of ordinoides for the reason that 

 it intergrades mth elegans throughout the entire length of its range." 

 From this two inferences may be drawn: These species are either 

 valid and the specimens are amenable to unquestionable separation, or 

 our recognition of them as two distinct forms must give way to the 

 fact that there is complete intergradation, hence they are not distinct 

 and require to be united. 



The initial step in the direction of attempting to ascertain the final 

 status of these two species was taken when 50 specimens of T. ordi- 

 noides from the Sausalito Peninsula were subjected to a critical 

 examination.^ The present contribution offers the data obtained 

 from a similar number of specimens captured in Golden Gate Park on 

 the San Francisco peninsula along ^vith a comparison of the two series. 

 As a further illustration of the need of these and similar studies there 

 is appended the record of 80 specimens from twelve different localities. 



METHODS. 



In order tbat the data be correctly coordinated it is necessary to 

 ascertain the highest scale row count that obtains in the genus. In 

 Tha^nnopTiis the maximum is 23 rows. The paired rows are desig- 

 nated by permanent numbers from I to XI and the median by M, the 

 count being made from without inward. 



When the number of scale rows is decreased the sequence of sup- 

 pression is constant and is as follows: 



23 rows, V row suppressed, leaving, 

 21 rows, VI row suppressed, leaving, 

 19 rows, IV row suppressed, leaving, 

 17 rows, VII row suppressed, lea^dng, 

 15 rows, which are continued to the vent. 



In dealing with a specimen in which the maximum count is 21 rows 

 it is necessary to regard the V row as suppressed constructively. In 

 enumerating the rows one must count I, II, III, IV (V suppressed), 

 VI, VII, VIII, IX, X, XI, and the median. In such a specimen when 

 the 21 rows are reduced posteriorly to 19 rows it wiU be found to be 

 due to the termination of the fifth, row in actual counting, but this 

 row in terms of the generic count is the VI row and must be so re- 

 corded. Wben the 19 rows are decreased to 17 it is due to the ending 

 of the fourth row in actual counting, which is also the IV row in the 

 generic sense. When the 17 rows are reduced to 15 it is due to the 

 loss of the fourth row in actual counting, which in this case is the VII 

 row in terms of the generic count. Constant attention to these 

 details is imperative. 



1 1908, Bull. 61, U. S. National Museum, p. 1-49. 



2 1914, Thompson, Proc. U. S. Nat. Mus., vol. -IT, pp. 351-360. 



