POISONOUS SNAKES OF NORTH AMERICA. 387 



viz, in the '' Columbian Magazine or Monthly Miscellany " for November, 

 1786, as follows : 



The common number of libuke seldom exceeds 14 or 15 in a rattle; but tlie cue 

 given (fig. 4) is certainly a very great curiosity, even to a person who has seen a 

 great number of this genus of snakes. The libuhi? are 44 in number. The snake 

 from which this rattle was taken was not, as might be expected, of a size propor- 

 tionate to the prodigious length of its rattle, but rather of a middling-sized snake. 

 It was killed some time in the summer of this year, at Fort Allen.* 



The most plausible explanation seems to be that some practical joker 

 joined several rattles, thus deceiving the unwary paragrapher.t The 

 greatest number of joints recorded so far by a fully trustworthy 

 authority is, I believe, 21 as seen by Holbrook (N. Am. Herpet., 1 ed., 

 II, p. 85), unless we accept Miss Hopley-s somewhat vague statement, 

 possibly only made on hearsay, that "a Crotalus [species not given] 

 at the London Reptilium had twenty -five links at one time" (Snakes, 

 p. 301), or Catesby's figure of a G. horrklus witli -1 joints. (Carolina, 

 1743, II, pi. xli.) 



If the reader will again examine the representation of a perfect and 

 full-grown rattle, given above (fig. 23), it will be seen that the basal 

 rings are nearly equal in size, while the terminal ones rapidly diminish 

 in size toward the tip. The latter were, of course, formed during the 

 early years of the snake's life, when its growth was rapid, the increas- 

 ing size of each succeeding joint testifying to the corresponding increase 

 of thickness of the end of the tail. Later on this increase is much 

 slower and the new joints do not materially differ in diameter from the 

 preceding one. If, therefore, in the adult snake the rattle breaks 

 near the base, the new rattle which will grow out to succeed the lost 

 one will have an entirely different shape, the upper and lower outlines 

 being nearly parallel and the terminal button as large as the basal 

 ring, while the perfect, unbroken rattle tapers off toward a small and 

 rather elongate button. This difference between the parallelogrammic 

 rattle and the tapering rattle was once held to indicate specific differ- 

 ences and was used as a systematic character, but Garman has effec- 

 tually disposed of this notion in the paper on the rattle so often referred 

 to above. 



When crawling over the ground the rattle, as shown in fig. 23, is 

 carried with the greatest width vertical and somewhat raised off' the 

 ground, so as to prevent it from dragging. 



It is often sounded even in this way, but the usual position assumed 

 by the snake when vibrating the rattle in earnest is when coiled up and 



"" Quoted from DeKay, Zool. N. Y., in, 1892, p. 56, footnote. Of course, we iilace no 

 more credence in the story of the man who told Kalm that in his younger days he 

 had killed a Rattler with 30 joints than in the one quoted above. (Kalm, Svenska 

 Vetensk. Acad. Handl., xiii, 1752, p. 316.) 



1A correspondent recently informed the Smithsonian Institution that lie was in 

 possession of a " fossil rattle " of about 45 joints. The fossil, when received, proved 

 to be an Orilioceras, a mollusk of the Devonian age. 



