With the publication of Daniel Keren, Jamie McCarthy, Harry W. Mazal’s article “The Ruins of the Gas Chambers: A Forensic Investigation of Crematoriums at Auschwitz I and Auschwitz-Birkenau” (Holocaust and Genocide Studies, 2004, vol. 18, no. 1) the slow-burning debate on the number of the Zyklon B introduction openings in the morgue of crematorium I in Auschwitz Stammlager moved forward. It was not the last word on the topic, but it was a very prominent contribution.
The authors introduced a novel hypothesis: contrary to all the (admittedly contradictory) testimonial evidence there had been 5 Zyklon B holes, as shown on the diagram below (copyright is the authors’):
![]()
Mazal et al. based their conclusion on the examination of a 1945 photo, in which the marks on the roof of the former gas chamber, now an air-raid shelter, seemingly correspond to the little chimneys we see today, as well as to one still sealed opening (Z1).
Not entering into the debate on the location of Z1 (there have been objections by the usual suspects), I should note that calling it a Zyklon B introduction hole is in any case premature. Assuming the hole existed when the morgue was used as a gas chamber, it is still not clear that it functioned as one of the introduction openings, rather than serving some other function. Also, when we examine its photograph (fig. 31 in Mazal et al.’s article) , it is not clear what shape it had – it could have been round, as well as square. And indeed, it was not even reconstructed by the Poles. Hence I, for one, cannot accept it as a proven Zyklon B opening. That leaves me with holes Z2-Z5.
Hans has previously gathered and analyzed several testimonies concerning the Krema I holes in this article, on which I rely here.
The largest number of holes - 6 - was testified about by Broad and Müller. Broad’s (a relatively “outside” observer’s) claim is in any case exaggerated, since there aren’t enough sealed holes of the same size and form to account for it. And Filip Müller’s testimony on this seems to have been influenced by Broad. Moreover, according to Müller’s testimony during the Frankfurt Auschwitz trial (and contrary to his later book) he spent just about 6 weeks in the Auschwitz I “Fischl-Kommando”, at the very beginning of his stay in Auschwitz. He was then moved to Buna, then to Birkenau, where he worked in the crematoria quite a bit more intensively, so his memories about that earlier period were probably not as reliable when it comes to such details.
Golik and Pys testified about 4 holes. It should be noted however, that there is a danger that late testimonies about 4 holes could have been influenced by the physical situation after the Polish reconstruction.
There are also several testimonies about there being two holes. I do not include among them the ones that merely say that two SS-men poured in Zyklon B pellets - this doesn’t, by itself, exclude a possibility of a higher number of openings themselves. Among those who explicitly testified about 2 holes were Stark and Jankowski. As Hans pointed out, both were in a good position to observe the number of holes. However, Stark could have been influenced in his testimony by the common procedure of 2 SS men pouring in Zyklon B, and this cannot be excluded for Jankowski too, who, although an observer from the inside of the gas chamber, may have had to see mostly one or two Zyklon B piles (as well as possibly one or two open chimneys), which could also have had influenced his testimony.
Then there’s Aumeier, who testified about “2-3” openings and who also was in a good position to know the number of holes.
Finally there’s Grabner’s 1945 statement in which he very explicitly mentions 3 square holes made on Fritzsch's initiative. Of all the above I would expect Grabner to make the most exact statement: as the head of the Political Department in the relevant period he was responsible for the crematorium (G. Morsch, "Organisations- und Verwaltungsstruktur der Konzentrationslager", in W. Benz, B. Distel (Hrsgb.), Der Ort des Terrors: Geschichte der nationalsozialistischen Konzentrationslager, 2005, Band 1, S.66; A. Lasik, "Struktura organizacyjna KL Auschwitz", in W. Długoborski, F. Piper (red.), Auschwitz 1940-1945. Węzłowe zagadnienia z dziejów obozu, 1995, tom I, s.129) and took active part in its technical nitty-gritty (as we know from his correspondence), as well as in the gassings there.
So far, the testimonies about 2, 3 or 4 holes are nevertheless compatible with there being 4 “reconstructed” chimneys today: of course, in the case of 2 or 3 holes one would have to conclude that the reconstruction was partially wrong. Which is not a far-fetched assumption, despite Zlobnicki's account, given the other mistakes the Poles made during the reconstruction. Whether in such a case unrelated but heretofore existent holes that had used to have some other function were used, or whether new holes were created, would have to remain unknown on present evidence.
In The beginnings of the extermination of Jews in KL Auschwitz in the light of the source materials, 2014, Bartosik et al. argue (pp. 27, 52) that the "luftdichte Klappen" for Krema I mentioned in a 25.09.41 work card are the covers for the 4 chimneys. But the authors never establish this, they merely claim it. We would not expect a fully separable cover to be called a "Klappe", which is something that is usually fastened on one end:
Now let’s compare the position of the 4 existing chimneys (Z2-Z5) with the original situation in the crematorium before the reconstruction:
![]()
Before I offer my short analysis, I should warn that it’s a should-woulda-coulda type of argument that one would not normally use and that, juxtaposed with “hard” evidence, cannot beat it. But exactly given the scarcity and uncertainty of evidence such arguments can play a role in calculating the “probability estimates”.
It would seem that the 4-holes hypothesis is the weakest one because the position of Z3 just doesn’t make that much sense. It is true that one could think up an ad hoc hypothesis to explain this (e.g. the people at the door should have died first), but prima facie its placement speaks against it being one of the original holes (unless, as mentioned, its authenticity is confirmed by really clear and good evidence - which we don't have in this case).
We are left with 2 or 3 holes. But no 2-holes combination seems to be all that rational either. Z4 and Z5 are on one side of the roof. And Z2&Z4 or Z2&Z5 don’t make for a good combination for a proper distribution of Zyklon B pellets. The most symmetrical solution that would have provided good gas distribution seems thus to be Z5&Z2&Z4.
To sum up: each of the above combinations is possible, but the one that makes most sense is the one with 3 holes. This also corresponds to the testimony of Grabner, the witness who was in the best position to know this detail.
I would therefore posit that originally there were 3 holes in the roof of Krema I. Of course, given the nature of the sources, this will have to remain a conjecture - until new evidence either confirms or refutes it.
The authors introduced a novel hypothesis: contrary to all the (admittedly contradictory) testimonial evidence there had been 5 Zyklon B holes, as shown on the diagram below (copyright is the authors’):
Mazal et al. based their conclusion on the examination of a 1945 photo, in which the marks on the roof of the former gas chamber, now an air-raid shelter, seemingly correspond to the little chimneys we see today, as well as to one still sealed opening (Z1).
Not entering into the debate on the location of Z1 (there have been objections by the usual suspects), I should note that calling it a Zyklon B introduction hole is in any case premature. Assuming the hole existed when the morgue was used as a gas chamber, it is still not clear that it functioned as one of the introduction openings, rather than serving some other function. Also, when we examine its photograph (fig. 31 in Mazal et al.’s article) , it is not clear what shape it had – it could have been round, as well as square. And indeed, it was not even reconstructed by the Poles. Hence I, for one, cannot accept it as a proven Zyklon B opening. That leaves me with holes Z2-Z5.
Hans has previously gathered and analyzed several testimonies concerning the Krema I holes in this article, on which I rely here.
The largest number of holes - 6 - was testified about by Broad and Müller. Broad’s (a relatively “outside” observer’s) claim is in any case exaggerated, since there aren’t enough sealed holes of the same size and form to account for it. And Filip Müller’s testimony on this seems to have been influenced by Broad. Moreover, according to Müller’s testimony during the Frankfurt Auschwitz trial (and contrary to his later book) he spent just about 6 weeks in the Auschwitz I “Fischl-Kommando”, at the very beginning of his stay in Auschwitz. He was then moved to Buna, then to Birkenau, where he worked in the crematoria quite a bit more intensively, so his memories about that earlier period were probably not as reliable when it comes to such details.
Golik and Pys testified about 4 holes. It should be noted however, that there is a danger that late testimonies about 4 holes could have been influenced by the physical situation after the Polish reconstruction.
There are also several testimonies about there being two holes. I do not include among them the ones that merely say that two SS-men poured in Zyklon B pellets - this doesn’t, by itself, exclude a possibility of a higher number of openings themselves. Among those who explicitly testified about 2 holes were Stark and Jankowski. As Hans pointed out, both were in a good position to observe the number of holes. However, Stark could have been influenced in his testimony by the common procedure of 2 SS men pouring in Zyklon B, and this cannot be excluded for Jankowski too, who, although an observer from the inside of the gas chamber, may have had to see mostly one or two Zyklon B piles (as well as possibly one or two open chimneys), which could also have had influenced his testimony.
Then there’s Aumeier, who testified about “2-3” openings and who also was in a good position to know the number of holes.
Finally there’s Grabner’s 1945 statement in which he very explicitly mentions 3 square holes made on Fritzsch's initiative. Of all the above I would expect Grabner to make the most exact statement: as the head of the Political Department in the relevant period he was responsible for the crematorium (G. Morsch, "Organisations- und Verwaltungsstruktur der Konzentrationslager", in W. Benz, B. Distel (Hrsgb.), Der Ort des Terrors: Geschichte der nationalsozialistischen Konzentrationslager, 2005, Band 1, S.66; A. Lasik, "Struktura organizacyjna KL Auschwitz", in W. Długoborski, F. Piper (red.), Auschwitz 1940-1945. Węzłowe zagadnienia z dziejów obozu, 1995, tom I, s.129) and took active part in its technical nitty-gritty (as we know from his correspondence), as well as in the gassings there.
So far, the testimonies about 2, 3 or 4 holes are nevertheless compatible with there being 4 “reconstructed” chimneys today: of course, in the case of 2 or 3 holes one would have to conclude that the reconstruction was partially wrong. Which is not a far-fetched assumption, despite Zlobnicki's account, given the other mistakes the Poles made during the reconstruction. Whether in such a case unrelated but heretofore existent holes that had used to have some other function were used, or whether new holes were created, would have to remain unknown on present evidence.
In The beginnings of the extermination of Jews in KL Auschwitz in the light of the source materials, 2014, Bartosik et al. argue (pp. 27, 52) that the "luftdichte Klappen" for Krema I mentioned in a 25.09.41 work card are the covers for the 4 chimneys. But the authors never establish this, they merely claim it. We would not expect a fully separable cover to be called a "Klappe", which is something that is usually fastened on one end:
[Klappe:] bewegliche Vorrichtung zum Schließen einer Öffnung; Gegenstand, mit dem sich etwas verdecken, auf- und zumachen lässtIn this case a Klappe can easily be a ventilation flap. Which is confirmed by the fact that it was to be produced from black iron plate, as is stated on the reverse of the document (published in C.Mattogno, Auschwitz: Crematorium I, 2005, p.121). Hence the document cannot serve as evidence in the issue of the number of Zyklon B openings.
[klappen:] etwas, was mit etwas auf einer Seite verbunden ist, in eine bestimmte Richtung bewegen
Now let’s compare the position of the 4 existing chimneys (Z2-Z5) with the original situation in the crematorium before the reconstruction:
Before I offer my short analysis, I should warn that it’s a should-woulda-coulda type of argument that one would not normally use and that, juxtaposed with “hard” evidence, cannot beat it. But exactly given the scarcity and uncertainty of evidence such arguments can play a role in calculating the “probability estimates”.
It would seem that the 4-holes hypothesis is the weakest one because the position of Z3 just doesn’t make that much sense. It is true that one could think up an ad hoc hypothesis to explain this (e.g. the people at the door should have died first), but prima facie its placement speaks against it being one of the original holes (unless, as mentioned, its authenticity is confirmed by really clear and good evidence - which we don't have in this case).
We are left with 2 or 3 holes. But no 2-holes combination seems to be all that rational either. Z4 and Z5 are on one side of the roof. And Z2&Z4 or Z2&Z5 don’t make for a good combination for a proper distribution of Zyklon B pellets. The most symmetrical solution that would have provided good gas distribution seems thus to be Z5&Z2&Z4.
To sum up: each of the above combinations is possible, but the one that makes most sense is the one with 3 holes. This also corresponds to the testimony of Grabner, the witness who was in the best position to know this detail.
I would therefore posit that originally there were 3 holes in the roof of Krema I. Of course, given the nature of the sources, this will have to remain a conjecture - until new evidence either confirms or refutes it.