Education

Staff Publications / Articles

 

Irradiation 101

Morning Glory Goblet & Wine

The Chain-Border Pressed Glass Tray 

 

---

 Irradiation 101

This article is an overview of a lecture that I presented at the 2000 Early American Pattern Glass Society convention, along with the results of ongoing tests by myself and glass artist Art Reed. In 2000 we were aware of the practice of subjecting colorless pressed glass to radiation with the intent to alter the glass to a desirable and more valuable color; the dilemma we faced was how to detect this unscrupulous practice.

The history of irradiating glass goes back to at least the mid 1970s. At that time the process was used, primarily by the glass industry of the Czech Republic, as a cheaper method of producing color, albeit in a very limited range. The first dramatic impact on the antiques world occurred in the late 1980s/early 1990s when examples of Lalique glass in a previously unrecorded deep purple began to appear. Single pieces sold at auction for as much as $50,000 before suspicions began to arise which ultimately resulted in fraud charges being leveled against a London dealer. Now all pieces of deep purple Lalique are considered extremely suspect, including all of the previously published examples. In the mid 1990s news of fruit jars and insulators surfacing in radiation-induced colors caused extreme concern in both collecting fields.

Irradiation can be accomplished at any facility with gamma ray or electron beam capabilities. This includes commercial sterilization plants that serve the health care and food industries, along with research facilities, many of which are affiliated with colleges and universities. Both methods disrupt the atomic structure of the glass which affects the atoms’ ability to transmit light, ultimately changing the color of the glass. At the present time, the composition of the glass (basically the decolorization agent) appears to be the primary factor as to the resulting color. In testing both colorless lead and soda-lime formulas, the only colors achieved were in the purple and brown spectrums.

Keeping in mind that we are only dealing with American pressed glass produced between approximately 1830 and 1920, we will evaluate the resulting colors. The irradiated browns range from a light smokey topaz to a darker brown, but not a true amber. Dr. Robert Brill of the Corning Museum of Glass suggests that these browns are the result of either cerium oxide or selenium being used as the decolorization agent. Obviously these are not colors used in the medium or period that we are concerned with here; they do however cause concern in the fruit jar and insulator fields.

The irradiated purples, which range from a lavender to a deep purple, occur due to 19th century glassmakers’ use of manganese in small amounts to decolor the batch. The same glass manufacturers also added manganese in larger amounts to obtain a purple color (a distinctly different color than the amethyst of the same period which has a reddish shade). Unfortunately, the antique purple and the "new" purple are nearly indistinguishable. In the limited number of examples that I have compared, the irradiated purple has a slightly bluer tone than the old purple. The similarity of these two colors is quite troubling, especially to collectors of lead based wares of the 1830s to 1860s.

Facts relating to the detection and reversal of the results of irradiation have been cloudy at best, with rumors being most often encountered. We have found no simple test to identify an irradiated piece of glass. A Geiger counter is of no use because the glass retains no radioactivity. Similarly, neither short-wave nor long-wave UV (black) light has provided any assistance. To date, the only proven test known to us is actual reversal. 

Art Reed has been successful in converting irradiated glass back to colorless, albeit in limited but ongoing tests. He has accomplished this by slowly reheating the glass in an annealing oven, which causes the manganese atoms to revert to their original structure, thereby returning the glass to its colorless state. One of the dangers of this procedure is that different periods of glass appear to have differing temperature points of reversion and melt down. The lead-based examples that we have tested reverted at between 600 and 700 degrees Fahrenheit while the soda-lime example, which was actually the product of natural solarization (sun-purpled), had to be heated to a dangerously high 1050 degrees. We have not yet obtained an irradiated example of soda-lime glass for testing, but hope to in the near future.

Art is currently performing "slump tests" on glass of differing periods in hopes of compiling a maximum temperature scale which will minimize the chance of the test object softening or melting completely. We would like to strongly caution novices against using this reversal process because of the high risk to the glass itself. If an article of glass is heated or cooled too quickly, it will crack or explode from thermal shock. Furthermore, if a piece is overheated it will, and I quote Art, "slump into a puddle."

Our research up to this point has been very encouraging. While purple-irradiated lead formula glass presents a tricky obstacle for collectors and dealers, thankfully there is a foolproof test that can be used to prove or disprove the authenticity of the color. On the other hand, testing and reverting soda-lime formula glass that has been sun-purpled or irradiated appears to be more problematic due to the apparent need for much higher temperatures. A great deal more testing is needed in order to determine the feasibility of using this process safely on pressed glass of this later period.              

---

Morning Glory Goblet & Wine:

Originals versus Reproductions

1.  Examine all four vines running up the stem on the mold lines. On the original, all the vines have a stem branching to the left near the bottom of the bowl. On the reproduction, one vine has no stem branching to the left at all. That stem is the one that has the unattached floating leaves above it, as described in point #3.

2.  Find the open bloom on the lower part of the bowl that points to the right. On the reproduction, the center stamen ball on this bloom is much larger than the other four balls. On the original, all of these balls are about the same size.

3.  This same bloom on the original has a stem passing behind it (just under the lower stamen) which connects the leaf to the left of the bloom to the main stem on the right. On the reproduction, this leaf is just floating with its stem going into the bloom but not coming out.

4.  Also, on the reproduction the stem above this bloom is not connected to anything, thusly all of the leaves above it are just floating with no connecting stem to the main vine.

5.  Move right one seam line on the pattern. The leaf pointing down on this vine has two stems on the original and only one on the reproduction.

6.  Move right one seam line on the pattern. Notice at the top of the goblet the two horizontal leaves with downward pointing leaf between. On the original, the right horizontal leaf has a stem that connects to the main vine. On the reproduction, the right horizontal leaf has a stem that connects to the top of the downward pointing leaf only.

7.  Move right one seam line on the pattern. Notice the open bloom which is pointing up on this seam line. The reproduction has five distinct stamen balls on the top edge outside of the bloom. The original has only four balls on the outside edge, with the right hand ball actually inside the bloom, plus another barely noticeable ball directly on top of the mold seam.

1.  Find the open bloom on the lower part of the bowl facing left. On the original, the center stamen ball touches the leaf to the left and the upper ball does not touch the leaf above it. On the reproduction, the center ball does not touch the leaf to the left and the upper ball touches the very edge of the leaf above it.

2.  Also on the reproduction, the group of four leaves and one closed bloom to the left of the aforementioned open bloom does not have a connecting stem to the main vine below.  The original has the connecting stem on the top of the mold line.

(back to top of article)   (back to top of page) 

---

The Chain-Border Pressed Glass Tray

The pressed chain-border tray[1] was arguably the most ambitious and complicated form produced or attempted by the Boston and Sandwich Glass Company or any of their competitors during the early period of American pressed glass. Machine pressing of glass is generally accepted as being invented in the United States in either 1825 or 1826.[2] The initial production consisted of small simple articles like bases for free-blown lamps, as well as candlesticks, cup plates, salts, and furniture knobs. Most pieces of the earliest period featured primarily angular ornamentation combined with simple scrolls, hearts, and fans.[3]

As mold makers and glass workers become more comfortable with the new concept of pressing glass, technological improvements like the cap ring, allowed for the manufacturing of wares with more intricate and stylistically pleasing designs.[4]  These improvements would usher in what is considered by many to be the most aesthetically beautiful and technically important period of pressed glass in America, the so-called “Lacy Period” of 1830-1860.

The subject of this article explicates the difficulties that mold makers and glass workers encountered as they explored the limitations of the new pressing technology while it was still in its infancy. The exact production period of the chain-border tray is undocumented, but it is thought to date as early as 1830 and possible as late as 1845.[5]  It is firmly attributed to the Boston and Sandwich Glass Company based on fragments dug at the factory site-first reported by Dr. Charles W. Green in 1937[6] and most recently corroborated by Raymond arlow & Joan Kaiser.[7]

The tray has been recorded in three distinct variations, each featuring the open-chain border and open handles, as well as a newly discovered variation wit neither the border nor the handles. The three chain-border variants were first discussed by Edith Gaines in a 1971 article in The Magazine Antiques.[8]  The plain-rim variation was discovered by the author in 2006. All examples feature an identical complex central design of diamonds, scrolls, lilies and rosettes, all surrounded by a rope-like table ring. The variations, which occur in the rim and shoulder design, appear to be the result of the need to rework the original mold due to a deficiency and/or breach to the mold itself.

The first version of the tray (fig. 1), produced from what appears to be the original mold, features a rope-like top edge on the chain border and four 10-point stars on the inner border. These stars are placed symmetrically on each side and end within each of the four peacock-eye plumes. The singular example of this tray recorded originally in the William J. Elsholz collection[9] had been broken and repaired with staples. 

In the second version of the tray (fig. 2), the rope-like top edge has been erased leaving the upper portion of the chain unpatterned. This modifications appears to be the initial attempt to improve the efficiency of the mold. The original serrated design on the top of the rim would have made it more difficult for the molten glass to fill in the chain completely[10] as well as hindering removal from the mold after pressing. The only known example of this variation, also from the Elsholz collection[11] exhibits a small unintentional lump of glass on the reverse of the shoulder at one of the stars – apparently made by a seepage of molten glass through a small mold break. It seems reasonable to theorize that this break was the result of the machining process needed to erase the rope design on the upper rim. 

The third, and most often encountered version[12] exhibits only two of the original four inner-border stars, one at each end below the handle. In this version the rim remains the same, but the mold was rather crudely repaired at the aforementioned shoulder break, which resulted in the elimination of the corresponding star. Presumably in order to keep the overall design symmetrical, the star on the opposite shoulder was removed in the same manner. On all of the examples of this version that I have observed, the stippling of the patched area is much less defined and the outline of the patch is quite obvious.

The fourth, and most likely final version of the tray, was produced without the chain edges and the end handles. The repaired mold used in version three was reworked yet a third time by removing all traces of the original open-work border and handles. At first glance, the author and numerous other glass scholars mistook this variation for an example with its rim and handles removed by post-production grinding and polishing, but upon close examination it became evident that this fourth variation tray exhibits a mold line encircling its rim indicating that it was pressed in the manner.[13]  There are only two examples of this borderless version tray currently known: the first example discovered by the author,[14] and a second example contemporaneously discovered by Jane Spillman, in the collection of the Corning Museum of Glass. It is very likely that additional examples of version one, two, and four or this tray will come to light as a result of this article.

The Boston & Sandwich chain-border tray represents an important achievement for American mold and glassmakers in the second quarter of the 19th century.   Unfortunately, there is very little information available concerning mold makers of this period, but it is quite obvious that the mold for this tray would have involved countless hours of tedious work to complete.[15]  Due to the unique open-work characteristics of the design, the glass factory pressman and workers had to overcome numerous production difficulties not previously encountered. The overall rarity of the tray in any of its variations would seem to indicate a short production period and/or a limited success rate of producing acceptable specimens. The extreme rarity of versions one and two indicates that the original mold and subsequent altered mold must have been in use for a very short period, possibly only a few turns for each. The fact that version three is the most commonly encountered variation suggests that the second mold modifications resulted in a somewhat more successful production period.

The extreme mold alterations of the fourth version, which completely changed not only the aesthetic integrity of the tray, but probably more significantly, the practical usefulness of the form, appear to be a final attempt by Sandwich or possibly a second firm to produce a satisfactory product from the original mold. Its rarity indicates that this final venture was unsuccessful and the mold was most likely abandoned, thus bringing to an end the saga of arguably the most arduous project undertaken by the early mold makers and glass pressers of early Industrialized America.

The author would be very pleased to learn of any new discoveries of versions one, two, or four of this tray.   

1.  Due to the absence of original source materials, e.g. company catalogs, which specifically names this particular form, the author has chosen the most commonly used designation, “tray” for means of brevity. Other authors have chosen to refer to this form as dish, a cake plate and a cake tray

2.  Kenneth M. Wilson, American Glass, 1760-1930: The Toledo Museum of Art. New York: Hudson Hills Press, 1994, p.265.

3.  For examples, see Raymond E. Barlow & Joan E. Kaiser, The Glass Industry in Sandwich, Volume 1. Atglen, PA: Schiffer Publishing, 1993, pp. 105-107, figs. 1001-1008; and Wilson [note 2], pp. 299-310.

4.  For discussions on early machine pressing and its development see Barlow and Kaiser [note 3], pp. 91-96; and Wilson [note 2], pp. 265-271. 

5.  Barlow and Kaiser place the tray’s production period as 1830-1840 ([note 3], p. 119, fig. 1047), while Wilson narrows the period to 1835-1840 ([note 2], p. 370, fig. 497); and Jane Spillman (American and European Pressed Glass in The Corning Museum of Glass, p. 101, fig. 300) expands the date to ca. 1830-1845. 

6.  Dr. Charles W. Green, “A Most Important Discovery at Sandwich,” The Magazine Antiques, v. XXXII, no. 2, August 1937, pp. 58-59.

7.  Barlow and Kaiser [note 3], p. 119, fig. 1047

8.  Edith Gaines, “The chain border tray: three versions,” The Magazine Antiques, v. C., no. 2, August 1971, pp. 256-257. The article was based on information supplied by James H. Rose and illustrations supplied by William J. Elsholz. 

9.  Richard A. Bourne Co. Inc., Public Auction: The Elsholz Collection of Early American Glass, Sessions V & VI, 1987, pp. 473-474, lot #1773. Elsholz acquired this tray from George and Michael Abraham, 5-27-48.

10.  One handle of the tray in Figure 2 is incomplete due to the molten glass not correctly filling in the mold. The handles of the tray in Figure 1 show a crack-like line in each handle due to the molten metal not properly melding. 

11.  Bourne, Elsholz, Sessions I & II, 1986, p. 131, lot #517. Elsholz acquired this tray from James H. Rose, 7-1-67, formerly in the collection of William Tallman Russel.

12.  This third variation is represented in numerable museum collections including The Sandwich Glass Museum, The Corning Museum of Glass, and the Toledo Museum of Art, as well as numerous private collections.  

13.  A number of altered examples of the third version of this tray exist that have had their damaged chain-border and sometimes handles removed by grinding and polishing. These post-production altered examples are easily identified because they show clear evidence of polishing to the rim and exhibit no indication of a mold line at the rim edge. 

14.  Green Valley Auctions, Inc., Seventh Annual Spring Cataloged Auction of Glass & Lighting, 2007, p. 122, lot #842. This tray was formerly in the collection of Bill Graham. 

15.  See Barlow and Kaiser [note 3], pp. 94-95, for a discussion of the relationship between Enoch S. Dillaway, Boston MA (a brass founder and mold maker) and Deming Jarves of the Boston & Sandwich Glass Company. 

(back to top of article)   (back to top of page)