Imagine that… a page of goat quotes!
1. May 26, 2006: light microscopy test report by K.D. Langley Fiber Services, as requested by CCMI (PDF: excerpted from Exhibit 10 of complaint).
Sample (as identified in the test results and attachment): Debbie Bliss Cashmerino Chunky, labelled as containing 55/33/12 merino wool/microfiber/cashmere.
Result: “The sample contains Wool and Acrylic fibers. No Cashmere fibers were observed” (this was not a quantitative test).
After this test was conducted, there was a TNNA show where the cashmere content issue was a hot topic of conversation (Complaint, paragraph 45). When the story broke, I don’t think that the second page of that PDF linked above was included; the inference I draw from the complaint is that it was circulated by Cascade Yarns later (Complaint, paragraphs 49 and 57).
After TNNA, KFI’s counsel sent a letter to Cascade Yarns alleging that Cascade Yarns had made “false and defamatory statements” at TNNA (PDF: excerpted from Exhibit 11 of complaint). This letter references KFI’s “independent lab reports analysing larger and more reliable samples” (not attached).
In response to this letter, Cascade Yarns’s counsel responded, requesting these test results (PDF: excerpted from Exhibit 11 of complaint). KFI’s counsel responded with a letter enclosing the lab reports (PDF: excerpted from Exhibit 11 of complaint (covering letter only)). And so we have:
2. January 2006: lab report from Laboratorio di Analisi Prove e Richerche Tessili (PDF: excerpted from Exhibit 11 of complaint).
Sample (as identified in the test results, from what I can make out in this grainy fax using the magic of online translators): a sample of 100% protein fibers.
Results: 55% cashmere, 45% wool or some similar fiber.
This appears to have been a microscopic examination in accordance with an ASTM standard. The requester looks like a Marcopolo srl.
Why is this sample a wool/cashmere blend only? Well, it might be a test of an input into a wool/acrylic/cashmere blend yarn, or it might be a test of the animal fiber component of an acrylic blend. However, we can’t tell from the document itself what was done with this fiber; if it were to be combined with acrylic microfiber to produce a yarn that was 33% acrylic, then 67% would have come from this wool/cashmere blend, meaning that the blended fibers would contain 36.85% cashmere — too much, unless cashmere fibers were lost in the manufacturing process. To arrive at the target cashmerino blend, you’d need 33% acrylic, 45.2% wool, and 21.8% of this blend (assuming I got that right), which seems like an odd way of going about things, but then again I know nothing about how the fibers are sourced by the mills.
(I don’t know, not being a spinner, of the likelihood of shorter-length fibers falling out of a properly blended top prior to spinning. As a knitter, I know that the shorter fibers are more likely to fall out with even casual use or pill on the surface of the garment with abrasion, but I’m relying on conventional wisdom and observation and not on an analysis of the contents of my vacuum cleaner bag; the latter would require that someone actually did the vacuuming around here.)
3. Rapporto di Prova (unknown date): page 2 of a lab report from “Laboratorio Accreditato No 0331 ITS” (PDF: excerpted from Exhibit 11 of complaint).
Sample (again, with difficulties in legibility and translation): tops containing wool, cashmere, and acrylic? The report looks like it has the word “Tops” (which makes me wonder if there is no Italian word for tops).
Results:: (1) 83.98% wool, 16.02% cashmere, as determined by microscopy at 500X and using some measurement software of the animal fiber part of the sample, apparently; (2) 61.7% animal fibers and 38.3% acrylic, as determined by a bleach test; with a conclusion of 51.8% wool, 38.3% acrylic, and 9.9% cashmere.
If this was meant to be a 55/33/12 wool/acrylic/cashmere yarn, the results are close; the cashmere is within the 3% tolerance, but the wool and acrylic are a bit off. However, we seem to be missing the first page of the report, which might have included identifying data such as the customer, the date, and the name of the lab.
4. July 7, 2006: report from TFT (Ilkley), Ltd., addressed to Designer Yarns (PDF: excerpted from Exhibit 11 of complaint).
Sample: “samples of ‘Debbie Bliss Cashmerino Aran’ yarn in a range of six shades”
Results: cashmere was found in all samples.
This was a qualitative analysis, not quantitative.
5. June 20, 2006: report from Quality Control Laboratory, for customer Designer Yarns (PDF: excerpted from Exhibit 11 of complaint).
Sample: three yarn samples, identified as “Brown 300008″, “Blue 300005″, and “Pink 300603″.
Results: From an examination of 1000 fibers, it was determined that the wool content was between 50.4 and 52.4% in all samples; the microfiber content was between 34.8% and 36.3%; and the cashmere between 11.9% and 13.3%.
However, TKW followed up with this tester; this report, apparently, was informal and done for a “friend of a friend”, and when the tester conducted the test she only looked at “three small wraps of yarn” (PDF: Exhibit 13 of complaint). And, in fact, she recommended Langley for testing in the United States. (Paragraphs 63 and 64 of the complaint contain the factual allegations detailing the “friend of a friend” chain.)
After these lab reports were exchanged, further tests were commissioned from Langley by TKW:
6. July 18, 2006: report from K.D. Langley Fiber Services for customer TKW (PDF: excerpted from Exhibit 9 of complaint).
Sample: three samples: Debbie Bliss Baby Cashmerino, KFI Cashmereno, and Debbie Bliss Cashmerino Aran
Results: From an examination of at least 500 fibers using light microscopy, no cashmere found in the samples.
7. July 25, 2006: report from K.D. Langley Fiber Services for customer TKW (PDF: excerpted from Exhibit 9 of complaint).
Sample: Debbie Bliss Cashmerino Aran.
Results: At least 500 fibers were examined using light microscopy, and a further test was conducted using a chemical method (essentially, the animal fibers are removed using bleach). No cashmere was found; moreover, the reported composition was 57.2% wool and 42.6% acrylic.
Another lab also carried out tests for TKW (with the same identifying lot numbers):
8. August 31, 2006: report from STR for customer TKW (PDF: Exhibit 15 of complaint).
Sample: three samples: Debbie Bliss Baby Cashmerino, KFI Cashmereno, and Debbie Bliss Cashmerino Aran
Results: No cashmere found in the samples.
This was a qualitative test only, using a microscopal method. The test report indicates that it had been “revised”: some kind of modification for fiber analysis and for overall rating was made to the original report of August 24.
Finally, KFI reported the results of a DNA test:
9. September 7, 2006: DNA test sent from KFI with a covering letter (I don’t have the actual test results, nor do I have a copy of the signed letter of compliance that was also attached). (PDF: Exhibit 21 of complaint)
The covering letter states:
These tests show categorically and unarguably that there is cashmere (goat) in Debbie Bliss Cashmerino yarns. It is not possible for the spinning mill to forget to put the cashmere in the blend. Making this yarn requires a very sophisticated process and unlike forgetting to put sugar in a cup of tea, it is impossible to forget to include a specific fiber. One thing for sure is that DNA testing does not lie. There is cashmere in Debbie Bliss Cashmerino.
No quantitative results are given. But sure, you can forget to put your sugar in! You taste the tea, make a face, then add your sugar. Unlike tea, however, when you discover that you forgot a fiber in your yarn, you can’t just spin that extra fiber in — you’d need to spin some more.
After test #1 was provided to KFI, KFI had this to say in a letter to its customers (which was also copied in a post on Knitter’s Review) (PDF: excerpted from Exhibit 11 of complaint):
… We have been informed that KD Langley always scores lower than TFT on “blind tests” that CMMI [sic] members receive from time to time to check their accuracy.
Third, as we have repeatedly pointed out to Cascade, it is difficult to test accurately for cashmere content, which can result in “false negative” results — i.e., findings of no cashmere content where cashmere in fact is present. Specifically, fiber experts we have contacted all state that, when one uses a projection microscope to examine Iranian or Mongolian Cashmere that is blended with extra fine merino wool — both of which have the same micron of approximately 18/19 — there will be some fibers that can be identified as wool, some as cashmere and others that are “indeterminate”.
CCMI does indeed hold round trials in which they send samples to their accredited labs for testing. The results of a 2005 trial are reported on CCMI’s website in this article, which also notes:
Wool may have been chemically altered to make it more difficult to distinguish from cashmere. Fine micron wool produced by a Chinese native sheep, which some in China have the unmitigated gall and effrontery to call sheep cashmere, is unfortunately painful and hard to distinguish by light microscopy. All of these products present challenges to the qualified laboratories able to identify these fibers.
(Look back at the stretched merino (in the second picture of that post) and compare them to the cashmere in the first picture.)
In the round trial reported here, some labs used light microscopy exclusively; others used scanning electron microscopy (think of those pretty Vogue Knitting pictures), BIO (I don’t know what that is), or DNA. CCMI states that in this trial of 28 labs, 24 used light microscopy (LM) exclusively:
Three labs using LM correctly identified all six samples. Four labs using LM correctly identified five of the six samples.
The four worst performing labs were one using LM that failed to correctly identify any of the samples and three using LM that each correctly identified only one of the six samples.
The results indicate that the “worst” labs used LM, and that the most difficult samples to analyze were the cashmere/wool blends since less than half of the labs got that right. However, that also tells us that some of the labs using LM — at least 8 of the labs using exclusively LM — must have correctly identified those blends.
KFI’s statement about Langley scoring “lower” is ambiguous — does this mean that given the same cashmere blend samples, Langley routinely finds a lower percentage of cashmere in the sample than TFT (which does not tell us which lab is more accurate)? Or is KFI suggesting that Langley does not correctly identify as many samples as TFT in the round trials? From the use of the word “score”, I guess they mean the latter.
However, even if you were to discount all of the microscopy tests, that probably leaves only tests #7 and #9. Test #7, which measured only the proportion of animal fibers to acrylic, does not rely on visual cues or magnification of fibers or the ability to distinguish cashmere from wool; and this test showed that there was nearly 10% acrylic than expected (these percentages are being expressed in absolute terms, obviously). And the excess amount of acrylic happens to be in the range we would expect for the cashmere (12%, plus or minus 3%). Thus, assuming these results are accurate, either the cashmere or wool content (or both) is less than the labelled quantity. Test #9 claims there is cashmere, but even if we assume this is accurate we do not know if this meets any legislated (after 2006) or accepted industry standard definition of cashmere (for example, if the micron count was too high), nor do we know how much cashmere was in the sample.
So, there we have it. We don’t need to accept anybody’s printed report (such as the ones we’ve seen here) as absolute truth without further investigation, but we’re not in a position to inquire into the conditions of each of these tests, and the origin of the samples provided. The parties will get to do that eventually.