majority belonged to the years l^etween aljout 1725 

 and 1750. Most of the bottles had the appearance 

 of being of English manufacture, although there is as 

 yet no method of identifying unmarked 18th-centurv 

 bottles of American colonial manufacture. A small 

 number of the bottles are of French origin, notably 

 the three ovoid flasks found in the primary deposit, 

 and others may be of Dutch or Rhenish manufacture. 

 Well represented were the bottles made specifically 

 as containers for Pyrmont mineral waters; five .seals 

 were found bearing this information. 



Only one personal seal was found in the pit; it 

 bears the initials "M. P." (Mann Page II) on a bottle 

 fragment of the 1760"s. This seal is surprisingly 

 simple. Most gentlemen of the 18th ccnturv pos- 

 sessed carefully executed bottle seals that often bore 

 their full names, dates, crests, and shields of arms or 

 rebus. It might therefore l)c expected that the Pages 

 would have possessed wine bottles bearing expensively 

 engraved seals befitting the cellars of so opulent a 

 mansion as Rosewell. That they did not has con- 

 sequently been construed as evidence of their penury. 

 However, the scarcity of sealed bottles is perhaps 

 a clearer indication of this than is the simplicity 

 of the seal, for there is ample precedent to show 

 that many colonial gentlemen used their tobacco 

 marks on their bottles, and we know that Mann 

 Page's mark comprised only his initials.'* 



Of greater interest in its own right is a seal (fig. 

 16. no. 1 ) that was unco\ered by ploughing in the field 



east of the mansion. This seal bears the initials 



and is clearly of the late 17th centur\-. The arrange- 

 ment of initials in the pyramidal form was generally 

 used to indicate a husband and wife combination, 

 the initials cf first names of the husband and wife 

 being capped by their surname initial. This arrange- 

 ment was accepted practice in England as early as 

 the 16th century, and it appears on thousands of 

 English wool bale seals in the 17th centurv; in the 

 second half of the same century it appears on many 

 beverage bottles made for taverns, indicating the 

 initials of both the licensee and his wife. In the 18th 

 century \'irginia planters and merchants often used 

 the triple initials as shipping marks. Howe\er, some 

 confusion creeps in when it is realized that these men 

 sometimes varied the long-established arrangement by 

 putting the initial of their middle name at the apex 



of the triangle. Thus, on the same page of a tea 

 account of 1769 we find the shipping marks " . for 



Robert and .\nn .\icholas and „ for Nathaniel L. 



Savage.^'' Further confusion resulted when .seme men 

 used different marks for the produce being .sent from 

 or to different plantations, the individual properties 

 being indicated by a symbol such as a diamond or 

 a mullet above two initials, or even an additional 

 identifying letter above them — thus creating again the 

 apparent triple initial triangle.^" 



The practice of making cheap bottle seals by cm- 

 ploying stock letters and setting them up in pairs 

 to order was common in the second half of the 17th 

 century but seems to have been rarely used in any 

 other period. Until the Rosewell seal was discovered 

 no example of the triple initial had ever been found 

 to have been set up in this way. For want of evidence 

 to the contrary, this seemingly unique seal is read 

 in the conventional manner, indicating perhaps some 

 such names as Thomas and Ann Osborne. 



Other glass items from the Rosewell pit included 

 fragments of square-sectioned bottles of the type fre- 

 quently identified as gin or case bottles. The illus- 

 trated section through one of the examples from the 

 pit shows that, in the absence of their necks, such 

 bottles could just as easily be called pickle jars (fig. 

 31, no. 13). Also present were fragments of large, 

 globular, thick-necked bottles: some of these fragments 

 probably came from wicker-encased carboys. Of 

 value as dating exidence were fragments from two 

 octagonal wine bottles (not illustrated) whose shapes 

 are comparable to examples bearing the name of 

 John Greenhow and dated 1770 that have been 

 found in Williamsburg excavations. 



The table glasswares from the pit are predominantly 

 of good quality and speak for themselves. However, 

 the straight-stemmed and trumpet-bowled example 

 from the primary filling (fig. 32, no. 7) is valuable as 

 daung evidence; it is of a type not in use prior to 

 about 1740. but this item probably dates somewhat 

 later. Other finds of table glassware included frag- 

 ments of an early lead glass decanter and pieces of two 

 rare bag-shaped cupping glasses. 



5' Morion Papers, op. ctl. (footnote 6), p. .^OO. 



'•' IhiJ., opposite p. 81. 



" C. Malcolm Watkins. "The Three-Inilial Cypher: 

 Exceptions to the Rule," Antiques, June 1958, vol. 73, 

 no. 6, pp. 564-565. 



PAPER 18: EXC.AV.ATIONS .\T ROSEWELL 



173 



