Ken Feder
Department of Anthropology
Central Connecticut State University
New Britain, CT 06050
E-mail: Feder@CCSU.CtStateU.edu

So, I will admit that it was with the usual trepidation that I returned the phone call of Mr. Ron Glazier who was anxious to know the significance of what he had found in his backyard. Ron had been given my name by a local car dealer who, himself, is a collector of New World antiquities. I must admit that I fully expected to find something along the lines of my petrified elephant's foot or meteorite fragment; I mean, after all, a blade cache with sixteen blades, one of which, according to Ron was more than seven inches in length, just seemed too good to be true, especially in someone's backyard, about five minutes from my house. Since it was so close, I decided to pay Ron a visit, fully prepared to disappoint yet another accidental archaeologist, having to explain for the hundredth time that, well, the basalt in the valley naturally breaks along straight cleavage planes and commonly forms long, thin, prismatic blade-like forms. "Yes, they do look very regular indeed, but they are entirely natural," I was ready to tell Ron. "How can you tell an artifact from one of these?" he would disappointedly likely ask me. "Well," I was prepared to respond, "You need to look for evidence of flaking, and look for bulbs of percussion," and on and on. You all know the drill.

There was only only problem with this nice, neat scenario I was prepared to play out. When I walked into Ron and Dottie's house, it was immediately obvious that Ron had found exactly what he said he had found; a remarkable cluster of sixteen enormous blades. Let me tell you that seeing these blades and the ensuing excavation makes up for all the petrified elephant feet, meteors, dreamt treasure, and giant bees I have ever or will ever be called on to investigate.

I must begin by recognizing the key role these incredibly generous people, Ron and Dottie, have played in all this by extending to us the privilege of excavating on their property and really opening up their home to us. Without Ron's curiousity and cooperation, without Ron and Dottie's great interest and recognition of the significant find on their land, and, it should be added, without the generous sharing of their bathroom facilities, none of this would have been possible.

Now, on to the blades. I first saw the collection in late 1991. As impressed as I was, I was unable to test the area around the place Ron described as the location where he found the blades because of scheduling, the weather, and, I will admit, a foolish assumption on my part. I figured that Ron had found all of the blades. I could only hope that perhaps some flakes or other artifacts would be found in the vicinity that might help us determine the cultural affiliation and age of the cache. It was not until the summer of 1993, with a large field school crew focusing on the McLean Game Refuge, that I had the time and a crew with sufficient expertise to investigate the area around where I presumed the complete cache had already been recovered.

In July and August of 1993, we laid in initially four and then one additional one-square-meter units in Ron and Dottie's backyard. Three of these units were located to the east, west, and north of the find spot as shown me by Ron. One of the units was located exactly over the find spot. The final unit excavated to date was later positioned immediately south of this unit.

As indicated, I was not terribly confident we would find anything. The area where Ron had found the points was pretty steeply sloping and I presumed that if there had been more blades, they had already been lost in the wetland at the base of the slope. I hoped that, perhaps, we would be lucky enough to find some debitage. If we were very lucky, perhaps we would find that the cache was part of a larger site on the property.

And, indeed, in three of the one-square-meter units we excavated initially, the situation was much as I expected it to be. Troweling down in three-centimeter increments and passing all matrix through one-eighth inch mesh hardware cloth, we found not so much as a flake in the three units located adjacent to where Ron had found the sixteen blades.

The unit situated directly over Ron's find spot was quite another story. Much to my surprise and delight, we soon determined that the cache had not been tapped out by Ron. Instead, fourteen more blades were recovered by my crew which included Barbara Calogero, Marc Banks, Kristen Janowski, Noel Coonce, Mike Zajko, Andrea Rand, and Pete Godwin.

These final fourteen of what grew to be, therefore, a thirty-blade cache were all located in situ; and at least nine of them were situated, as nearly as we can tell, in their original place of deposit. These nine blades had been neatly stacked in three layers in a shallow pit, overlapping and tightly clustered. From top to bottom, five blades were found, roughly oriented in a north-south line (some pointing north, others south), laid out flat on top of three more blades directly beneath them. Finally, one blade was located at the bottom of the entire pile. The five other blades we excavated were just a bit downslope from the tightly clustered nine just mentioned. These five appear to have washed out of the cache at some time in the past. There is no way of knowing, of course, if any other blades have washed out, down, and into the wetland.

The blades themselves are a remarkable grouping. They range in length from 11.7 to 18cm (4.6 to 7.1 inches). Their mean length is 13.6cm (5.4 inches). At their widest point from edge to edge they range from 3.9 to 5.7cm across (1.5 to 2.2 inches). Mean width is 4.63cm (1.8 inches). Maximum thickness measured from face to face near the base ranged from 1.0 to 1.3cm (.4 to .5 inches). Mean thickness was 1.16cm (.47 inches).

Click here to view the summary statistics and blade illustrations.

Summarizing those numbers briefly, the Glazier Blade Cache consisted of thirty extraordinarily long, for the most part narrow, and universally thin bifaces.

In terms of cultural affiliation and chronology, the cache is not a Late Archaic cremation burial. There is no evidence of burning. I believe that we can confidently state that the blade feature is not a Meadowwood cache. Joseph Granger (1981) has analyzed a large number of these caches. In his quite large sample, the mean length was only about 6.6cm, only about half the size of the blades in the Glazier cache. In fact, of the 518 blades Granger personally examined, the longest was shorter than the smallest Glazier blade by about 2.5cm (about an inch). Also, the Meadowood material tended to be proportionally broader; the mean width of 3.0cm is about 45% of the length of Meadowwood cache blades. At Glazier, mean width is only 34% of mean length. All this means is that the Glazier artifacts are proportionally longer and narrower than Meadowood blades.

The Glazier blades are not Snook Kill points. For comparison, the Waldron cache in South Windsor consisted of 24 Snook Kill points whose lengths ranged from 8.9 to 15.9cm. While these overlap somewhat with the Glazier blades, here again, they were proportionally wider, with some exceeding 7.6cm in width. Pretty clearly, Glazier is not a Snook Kill cache. The basic form and proportions of the blades are vaguely reminiscent of Paleo points, though none are fluted and none exhibit basal grinding or the typically concave bases of Paleo points especially in the Northeast.

The points exhibit some variability but, overall, remarkable consistency. A histogram of blade lengths shows a distribution fairly tightly clusterd about the mean, with a few long outliers in the distribution. A histogram of width shows much the same thing, with a tight clustering about the mean and slightly below, with a few wide specimens. A histogram of thickness has a quite different appearance, exhibiting an even tighter clustering about the mean, with few relatively thick or thin specimens.

For the more statistically inclined, if we graph the individual specimens by length, width, and thickness and compare them to the means and standard deviations, we again see a very tight clustering about the mean. When we do this for length we see a very small proportion fall outside of plus or minus one standard deviation from the mean. We see a less tight clustering for width, with a greater proportion falling outside of one standard deviation away from the mean. For thickness we see another very tight fit and an intersting difference; of those that fall more than one standard deviation away from the mean, almost all of them are more than one standard deviation below the mean.

Essentially, it seems clear that the manufacturer had a pretty good idea of what he or she wished to produceó and he did a damn good job of producing it! He or she wished to produce impressively long blades; perhaps the longer the better. Width was a less important variableóthat is, his or her tolerance seems to have been greater for blades that were wider or narrower than the standard. Finally, thin blades were desired; this variable was perhaps more important even than length. Again for the statistically minded, the coefficient of variation for thickness (7.4%) was lower than that determined for length or width, showing this very point. When we graph all three variables, length, width, and thickness, at the same time, we can present the result as a three-dimensional picture.

When we compare these variables one to another, we find less of a correlation than might have been expected. For example, graphing length against width shows no discenrable relationship; in other words, longer blades in the cache are not necessarily wider. It might have been thought that the necessities of the manufacturing process would have caused longer blades to be thicker. This turns out not to be true. When you graph length against thickness, there is only a vague positive correlation between the variables. Finally, there is even less of a correlation between width and thickness. This all merely bears out what I stated previously; the goal of the knapper seems to have been long, flat blades. The blades are all fairly narrow, but width seems to have been the least important consideration.

Identification of the raw material (or materials) is still unresolved. I can tell you that they are not flint, jasper, chert, chalcedony, quartz, or quartzite. They are far more likely a local material, conceivably a very fine grained basalt, or, perhaps a metamorphosed sandstone like hornfels. Barbara Calogero will be analyzing the raw material using x-ray diffraction.

Although weathering has taken its toll on the blade surfaces, flake scars are apparent on all the blades. We have x-rayed the entire assemblage, producing images where the flaking is far more obvious. Analysis of the flaking leads me to believe that the blades were nearly complete, end-stage blanks. I do not see any fine, tertiary flake scarring on any of the blades, though I am fully prepared to disagree with myself here.

Certainly much work remains to be done. In excavating the cache feature itself, and in the area around it, we were able to extract enough charcoal for at least a couple of radiocarbon dates. Unfortunately, erosion of the soil above the feature from runoff channeled over it by a culvert immediately to the north has brought modern debris into fairly close proximity to the cache. An iron nail was recovered in the soil just south of the intact portion of the cache, indicating other possible disturbance. We will be dating only that charcoal, therefore, that we are reasonably confident comes from among the bladesóand there is really not much of that.

We have not washed the fourteen blades found in situ. I will be contacting Tom Hester at the University of Texas about residue analysis.

So, obviously, the final chapter has not been written on this cache; to extend that metaphor we are barely into the first paragraph. Excavating the cache was exciting in and of itself. The analysis only makes it all the more exciting. I will be certain to keep you posted.