MUSEUM OF COMPARATIVE ZOOLOGY. 293 



Lyell shows a dike in his section of the Richmond coal field (c, 271 ; 

 our figure 17). Olmsted (c, 23G) and E. Mitchell speak of trap dikes 

 in North Carolina. Emmons does not describe or figure any sheets, but 

 shows several dikes on his sections {d ; our figure 24). Kerr states 

 that the sandstones are everywhere intersected in various directions by 

 dikes of trap ; their thickness varies from a few yards to two or three 

 rods, and their length occasionally reaches several miles : the sandstones 

 and shales are usually blackened for several feet or yards on either side 

 of the dilces (6, 14o) ; the dikes are usually transverse to the stratifi- 

 cation (a, 48). 



It would seem from this review that dikes are rare in Nova Scotia, 

 in Massachusetts and Northern Connecticut, and in New Jersey ; while 

 they are common in Southern Connecticut, in Pennsylvania, and farther 

 south. They ai'e of all sizes up to two hundred feet. Mill Rock and Pine 

 Rock just north of New Haven being the largest well-known examples. 

 They are, as a rule, dense and compact, with a rough columnar structure 

 at right angles to the sides ; some are reported as amygdaloidal, but 

 such are clearly exceptional ; their vesicles are very likely not of gas- 

 bubble origin. The sides of the dikes are not smooth, like the sides of 

 most of the dikes in the old jointed slates about Boston, but are more 

 or less irregular or even ragged (fig. 45) ; implying that there were no 

 joints to guide their fissures. It should be noted, however, that some 

 of the looser sandstone is still almost without joints, and in such cases 

 this point of evidence is of no value. Their metamorphic effects are not 

 far reaching so far as observed : a dike ten feet wide bakes the sandstone 

 for a foot ; a hundred-foot dike has some effect ten or twelve feet away. 

 None of these dikes have clearly the form of " necks " or " chimneys," 

 such as are described by A. Geikie about the Firth of Forth (Edinb. Roy. 

 Soc. Trans., XXIX., 1879, 4G8). 



Intruded Trap Sheets. — A number of sheets named in the following 

 list cannot be regarded as fully proven to be of intrusive origin : the evi- 

 dence for Pennsylvania and the States farther south is very incomplete. 



There seem to be no intruded sheets north of Middle Connecticut. 

 Farther south, near New Haven, East and West Rocks (G) and the 

 northern continuation of the latter have long been rightly regarded as 

 intrusions. Smaller examples occur at Wallingford (F), and doubtless 

 many more may be found. 



Percival's descriptions make it very probable that all his western line 

 of elevation from East and West Rocks at New Haven north to South- 

 ingtou are intrusive traps. In evidence of this, we may note the general 



