John Richardson

[Fleetwood Mac]

You can’t help yourself can you?

Never mind that. Where are your examples of experts who looked at Dr Phillips' situation and what did they say?

[Fleetwood Mac]

https://forum.casebook.org/forum/social-chat/creative-writing-and-expression/792993-on-trolls#post792993

I see.

You have no examples to back up your claim.

As expected.

[DJA]

As usual, you're mistaken.

Can I remind you that this is a forum to discuss the Whitechapel Murderer, as opposed to a forum discussing a fictional character who can't spell his own name.

[Abby Normal]

I'll tell you what it is, Sherlock, I'm curious about this one.

It seems that everyone on this thread, including those who support an earlier TOD, have accepted your proposition.

I'm not convinced and would like to hear more.

So, which experts? What did they say?

hey heres a thought. why dont you show a modicum of class and start addressing Herlock by his proper screen name.

[Doctored Whatsit]

He didn't. You've made this up.

Now this is getting really silly! Phillips made it crystal clear that his estimate could be incorrect when he said, "but it was right to mention that it was a fairly cool morning, and that the body would be more apt to cool rapidly from its having lost a great quantity of blood." That is a very obvious caveat.

[Doctored Whatsit]

He's quantifying time in terms of hours, i.e. "at least two hours".

So, when he says: "and probably more", it follows he is talking of hours.

You are fluent in English are you? "At least two hours and probably more", simply means a minimum of two hours and probably longer than that. Then he tells the coroner that this estimate might be wrong anyway.

[JeffHamm]

Hi all,

I've been having a bit more of a look at the cooling of internal body temperatures over time from the article I mentioned awhile back.

From looking at the limited data set, it generally looks like bodies cool following an exponential decay function until they reach the environmental temperatures. The drop in temperature from time A to time B will be larger the greater the difference between the body temperature and the environment. Except in the very early time period, where it seems it takes some time before a fairly stable cooling function forms.

I was wondering if the cooling rate (the constant k) is related to the body mass. And found that it appears to be, so if you know the BMI you can get a good estimation of that body's cooling rate. It also appears that the amount of time required to reach that stable cooling function is related to the BMI (I'll call that the "delay" period).

That got me thinking, it should be possible, using only the BMI, to produce a cooling function profile. Basically, estimate the k parameter from the BMI, and if the PMI for which you're estimating the temperature, just reduce k (cool a bit slower) by Time/Delay, so a simple linear function, where k just increases until it reaches stability.

Once you have those, if you have the environmental temperature, and set the initial body temperature to be 98.6 F, you should be able to plot the estimated temperature over time. That could then be used to produce an estimated ToD.

And while the results are certainly not better than current methods, I'm actually pretty pleased with how it's looking. The blue lines are the estimated cooling functions derived from the BMI, and the orange circles are actual temperatures at a known PMI. When the blue line is above the data, it means the PMI would be overestimated, and if the blue line is under the data, then the PMI is underestimated. Case 1 and 2 show quite a bit of overestimation (Case1 we're talking multiple hours, as in 3+ at times; Case 2 is over estimated by 1h 45min type thing; Case 4, underestimates by about 1h 45m at some of the longer PMI, etc). I've not fully calculated the error values, but for now I'm actually pretty pleased with the general fits.

What I now need to do is see if this works with a couple of cases that I've not used already as the equations I'm using are derived from these cases. It sort of has to work for this data set, the next step is to see if it produces a reasonable fit to cases not used to produce the BMI based model.

Anyway, with respect to Annie and Kate, given Kate was not as stout as Annie, she would be expected to cool faster.

- Jeff

[JeffHamm]

I've now been able to test the above with 3 other cases from the article (overlooked a figure with 3 more cases in it). In short, two of the cases end up producing roughly 2 hour type overestimations, while the 3rd is pretty good (inside of 30 minutes type thing); that's using the equations based upon the previous 6 cases, so given how unstable the equations should be due to the small number of cases, the fact that even the "misfits" are within +-3 hours, I'm pretty pleased with those outcomes.

Given the estimate the delay period part of this has been done very crudely (more or less look at some data I calculate and try and estimate where it appears to stop changing - I need to work out a more objective way to do that), I'm pretty pleased with how it is shaping up.

And to be clear, I'm under no illusions that this would result in a better method than is currently used, and even to get this to the best it can be would require a lot more than 9 cases (after having tested them, I can now add them to the data set and recalibrate the equations).

The next step will be to get a better profile of the error ranges. If I can get things to the point where I can automate a lot of this I might see if I can get my hands on the full data base of 85 cases. If I had that, I will be better able to see how well these relationships between BMI and the cooling constant, and the delay period, hold up. And, if they are reasonably strong, how well the estimated cooling profiles match with the actual cooling profiles.

That, however, is a story for another time.

- Jeff

[GBinOz]

Hi all,

I've been having a bit more of a look at the cooling of internal body temperatures over time from the article I mentioned awhile back.

From looking at the limited data set, it generally looks like bodies cool following an exponential decay function until they reach the environmental temperatures. The drop in temperature from time A to time B will be larger the greater the difference between the body temperature and the environment. Except in the very early time period, where it seems it takes some time before a fairly stable cooling function forms.

I was wondering if the cooling rate (the constant k) is related to the body mass. And found that it appears to be, so if you know the BMI you can get a good estimation of that body's cooling rate. It also appears that the amount of time required to reach that stable cooling function is related to the BMI (I'll call that the "delay" period).

That got me thinking, it should be possible, using only the BMI, to produce a cooling function profile. Basically, estimate the k parameter from the BMI, and if the PMI for which you're estimating the temperature, just reduce k (cool a bit slower) by Time/Delay, so a simple linear function, where k just increases until it reaches stability.

Once you have those, if you have the environmental temperature, and set the initial body temperature to be 98.6 F, you should be able to plot the estimated temperature over time. That could then be used to produce an estimated ToD.

And while the results are certainly not better than current methods, I'm actually pretty pleased with how it's looking. The blue lines are the estimated cooling functions derived from the BMI, and the orange circles are actual temperatures at a known PMI. When the blue line is above the data, it means the PMI would be overestimated, and if the blue line is under the data, then the PMI is underestimated. Case 1 and 2 show quite a bit of overestimation (Case1 we're talking multiple hours, as in 3+ at times; Case 2 is over estimated by 1h 45min type thing; Case 4, underestimates by about 1h 45m at some of the longer PMI, etc). I've not fully calculated the error values, but for now I'm actually pretty pleased with the general fits.

What I now need to do is see if this works with a couple of cases that I've not used already as the equations I'm using are derived from these cases. It sort of has to work for this data set, the next step is to see if it produces a reasonable fit to cases not used to produce the BMI based model.

Anyway, with respect to Annie and Kate, given Kate was not as stout as Annie, she would be expected to cool faster.

- Jeff

Hi Jeff,

An interesting project. It could get really complicated if you throw in Annie's malnourishment and her tuberculosis.

I have been pondering and researching your objective/subjective temperature synopsis, and it appears that the consensus of opinion is with subjective having been the case. While I accept that, my concern is still with the magnitude of the difference, so I would like to pose to you a few questions.

1. Would you consider it fair to say that the content of Phillip's caveat for Chapman, "but it was right to mention that it was a fairly cool morning, and that the body would be more apt to cool rapidly from its having lost a great quantity of blood.", would equally apply to Eddowes?

2. Brown stated that when he examined Eddowes body at the scene it was "quite warm" and estimated that she had been dead 30-40 minutes. If he had done another temperature feel test 30 minutes later, do you think it is possible that he would have said that she was "not quite cold"?

I accept the limitation of subjective temperature determination, but take into account that these procedures were being done by experienced police surgeons, not just GPs. While there is certainly a case for subjective errors, it is the magnitude involved that leaves me unpersuaded for the 1 hour PMI. I'm a little taken aback by the absolute opinions, replete with insults, being expressed on this thread (not by you), and find it interesting that the current poll on this site on the ToD for Chapman stands at 14 each, and on another thread luminaries such as Lynn Cates strongly supporting the earlier ToD and doubting the witnesses.

During my research I came across this, which may be of interest to those who have not closed their minds:

Echo, 19 Sep 1888:
Dr. G.B. Phillips, the divisional surgeon, has had another consultation with the police authorities respecting certain theories advanced. There are three points upon which there is agreement - that Annie Chapman was lying dead in the yard at 29 Hanbury street, when John Richardson sat on the steps to cut a piece of leather from his boot, his failure to notice the deceased being explained by the fact that the yard door, when opened, obstructed his view; that the poor creature was murdered in the yard, and not in a house, as had been at one time suggested; and that the person who committed the deed was a man with some knowledge of human or animal anatomy.

Publish date is 19 Sep. Richardson testified at the inquest on the 12th, Chandler and Phillips on the 13th, Phillips recalled, Long and Cadosch on the 19th. For the police to conclude that Richardson had missed the body due to the yard door, they must have done a re-enactment, possibly even involving Richardson, as he was considered a suspect.

For those who will decry the above as just a newspaper report, this is from Wolf Vanderlinden's excellent dissertation:

Swanson's report 19 Oct (Ref. HO 144/221/A49301C, ff. 137-45)
"If the evidence of Dr. Phillips is correct as to time of death, it is difficult to understand how it was that Richardson did not see the body when he went into the yard at 4:45 a.m. but as his clothes were examined, the house searched and his statement taken in which there was not a shred of evidence, suspicion could not rest upon him, although police specially directed their attention to him."

"Up to the present the combined result of those inquiries did not supply the police with the slightest clue to the murderer". "Again if the evidence of Mrs. Long is correct that she saw the deceased at 5:30 a.m. then the evidence of Dr. Phillips as to probable time of death is incorrect. He was called and saw the body at 6:20 a.m. [sic] and he then gives it as his opinion that death occurred about two hours earlier, viz: 4:20 a.m. hence the evidence of Mrs. Long which appeared to be so important to the Coroner, must be looked upon with some amount of doubt, which is to be regretted."

Home Office Files (Ref. HO/144/220/A49301C, f 8g)
" doubtful evidence points to some thing between 5:30 and 6: - but medical evidence says about 4 o'cl."

Far from "game over" I should think.

Best regards, George

[JeffHamm]

Hi Jeff,

An interesting project. It could get really complicated if you throw in Annie's malnourishment and her tuberculosis.

I have been pondering and researching your objective/subjective temperature synopsis, and it appears that the consensus of opinion is with subjective having been the case. While I accept that, my concern is still with the magnitude of the difference, so I would like to pose to you a few questions.

1. Would you consider it fair to say that the content of Phillip's caveat for Chapman, "but it was right to mention that it was a fairly cool morning, and that the body would be more apt to cool rapidly from its having lost a great quantity of blood.", would equally apply to Eddowes?

2. Brown stated that when he examined Eddowes body at the scene it was "quite warm" and estimated that she had been dead 30-40 minutes. If he had done another temperature feel test 30 minutes later, do you think it is possible that he would have said that she was "not quite cold"?

I accept the limitation of subjective temperature determination, but take into account that these procedures were being done by experienced police surgeons, not just GPs. While there is certainly a case for subjective errors, it is the magnitude involved that leaves me unpersuaded for the 1 hour PMI. I'm a little taken aback by the absolute opinions, replete with insults, being expressed on this thread (not by you), and find it interesting that the current poll on this site on the ToD for Chapman stands at 14 each, and on another thread luminaries such as Lynn Cates strongly supporting the earlier ToD and doubting the witnesses.

During my research I came across this, which may be of interest to those who have not closed their minds:

Echo, 19 Sep 1888:
Dr. G.B. Phillips, the divisional surgeon, has had another consultation with the police authorities respecting certain theories advanced. There are three points upon which there is agreement - that Annie Chapman was lying dead in the yard at 29 Hanbury street, when John Richardson sat on the steps to cut a piece of leather from his boot, his failure to notice the deceased being explained by the fact that the yard door, when opened, obstructed his view; that the poor creature was murdered in the yard, and not in a house, as had been at one time suggested; and that the person who committed the deed was a man with some knowledge of human or animal anatomy.

Publish date is 19 Sep. Richardson testified at the inquest on the 12th, Chandler and Phillips on the 13th, Phillips recalled, Long and Cadosch on the 19th. For the police to conclude that Richardson had missed the body due to the yard door, they must have done a re-enactment, possibly even involving Richardson, as he was considered a suspect.

For those who will decry the above as just a newspaper report, this is from Wolf Vanderlinden's excellent dissertation:

Swanson's report 19 Oct (Ref. HO 144/221/A49301C, ff. 137-45)
"If the evidence of Dr. Phillips is correct as to time of death, it is difficult to understand how it was that Richardson did not see the body when he went into the yard at 4:45 a.m. but as his clothes were examined, the house searched and his statement taken in which there was not a shred of evidence, suspicion could not rest upon him, although police specially directed their attention to him."

"Up to the present the combined result of those inquiries did not supply the police with the slightest clue to the murderer". "Again if the evidence of Mrs. Long is correct that she saw the deceased at 5:30 a.m. then the evidence of Dr. Phillips as to probable time of death is incorrect. He was called and saw the body at 6:20 a.m. [sic] and he then gives it as his opinion that death occurred about two hours earlier, viz: 4:20 a.m. hence the evidence of Mrs. Long which appeared to be so important to the Coroner, must be looked upon with some amount of doubt, which is to be regretted."

Home Office Files (Ref. HO/144/220/A49301C, f 8g)
" doubtful evidence points to some thing between 5:30 and 6: - but medical evidence says about 4 o'cl."

Far from "game over" I should think.

Best regards, George

Hi George,

I doubt anything I come up with will be of any direct application to the JtR murders because in order to use what I find we would need to have a stated rectal temperature reading. Even then, all of the data I've got access to so far is from intact bodies. The estimation of the cooling rate constant (k) and the initial "delay period" where k ramps up to the stable value, will produce values that I would be shocked to find are applicable to the types of murders we're dealing with. I would expect k to be larger than what gets estimated, for at least two obvious reasons. First, with the abdomens opened up, the cold air will get inside the body cavity, and therefore cool the inside of the body faster. Second, with the loss of blood involved, that's a large heat sink that's been removed, so again, the body should cool faster. But I have no information to base what the change in the k value would be, or how much that would shorten the delay period. As a result, these equations are not going to apply to the JtR murders even if we had a temperature reading.

As to your 1st question, I suppose it would. We know bodies will cool faster the colder the surrounding temperature. But again, Eddowes "felt warm", so examination by touch of the surface skin is being reported. Annie appears to have been walking around much of the night, Eddowes had been indoors (in jail, but inside nonetheless) until about 1 hour prior to her death. Annie's surface temperature at her skin could very well have cooled somewhat even before she was killed, meaning she may be starting at a lower temperature than Kate in the first place.

As to your 2nd, all I can say is perhaps. We know she would have cooled over that 30 minutes, but we would have to take into account the different amounts of clothing worn by the two victims. More clothes would be less cooling as they would trap the heat. There are a lot of things to consider, much of which we don't know the answers to, hence "perhaps".

And yes, the police were wrestling with the same questions we are today. We, at least, have access to more information about the reliability of both the medical evidence and witness statements. We know both are associated with more error than perhaps the police of the day realised (though they no doubt realised the witness statements can be suspect, they may have grossly underestimated how much error is associated with the medical).

- Jeff

[Fleetwood Mac]

Phillips made it crystal clear that his estimate could be incorrect when he said, "but it was right to mention that it was a fairly cool morning, and that the body would be more apt to cool rapidly from its having lost a great quantity of blood." That is a very obvious caveat.

Aye, the 'probably more' part of his statement.

Not the "at least two hours" part. The minimum time possible, his caveat did not relate to this part of his statement.

[Fleetwood Mac]

You are fluent in English are you? "At least two hours and probably more", simply means a minimum of two hours and probably longer than that. Then he tells the coroner that this estimate might be wrong anyway.

'One inconvenient problem with this. Dr Phillips didn't state: "at least two hours and probably longer than that". You're bending his words.

[GBinOz]

Hi George,

I doubt anything I come up with will be of any direct application to the JtR murders because in order to use what I find we would need to have a stated rectal temperature reading. Even then, all of the data I've got access to so far is from intact bodies. The estimation of the cooling rate constant (k) and the initial "delay period" where k ramps up to the stable value, will produce values that I would be shocked to find are applicable to the types of murders we're dealing with. I would expect k to be larger than what gets estimated, for at least two obvious reasons. First, with the abdomens opened up, the cold air will get inside the body cavity, and therefore cool the inside of the body faster. Second, with the loss of blood involved, that's a large heat sink that's been removed, so again, the body should cool faster. But I have no information to base what the change in the k value would be, or how much that would shorten the delay period. As a result, these equations are not going to apply to the JtR murders even if we had a temperature reading.

As to your 1st question, I suppose it would. We know bodies will cool faster the colder the surrounding temperature. But again, Eddowes "felt warm", so examination by touch of the surface skin is being reported. Annie appears to have been walking around much of the night, Eddowes had been indoors (in jail, but inside nonetheless) until about 1 hour prior to her death. Annie's surface temperature at her skin could very well have cooled somewhat even before she was killed, meaning she may be starting at a lower temperature than Kate in the first place.

As to your 2nd, all I can say is perhaps. We know she would have cooled over that 30 minutes, but we would have to take into account the different amounts of clothing worn by the two victims. More clothes would be less cooling as they would trap the heat. There are a lot of things to consider, much of which we don't know the answers to, hence "perhaps".

And yes, the police were wrestling with the same questions we are today. We, at least, have access to more information about the reliability of both the medical evidence and witness statements. We know both are associated with more error than perhaps the police of the day realised (though they no doubt realised the witness statements can be suspect, they may have grossly underestimated how much error is associated with the medical).

- Jeff

Hi Jeff,

For question 1, you are presuming a 5:30 ToD. Annie told her lodging house to hold her bed, so a late ToD would require Annie to be unsuccessfully plying her trade for 4 hours in an area where she was known without being seen by anyone willing to come forward to testify. An early ToD would put her closer to Eddowes condition. Would she have persisted so long before calling no joy and opting for rough sleeping somewhere like the Whitechapel graveyard?

For question 2, for your "perhaps" were to be accurate, wouldn't her perceived temperature then have to flat line for the next 12 hours?

While we do have the benefit of more knowledge, the police did have the advantage of experience, even though the profession was in it's infancy. While Trevor is often derided on this forum, I believe sometimes unfairly, he was a professional and speaks from an entirely different perspective to the layman.

Anyway, I'm delighted that we can discuss our differences of opinion without the adversarial "yes it is - no it isn't" repetition that is pervading this thread at present.

Best regards, George

[JeffHamm]

Hi George,

Hi Jeff,

For question 1, you are presuming a 5:30 ToD.

Yes, sorry, I thought you expected me to answer with that in mind.

Annie told her lodging house to hold her bed, so a late ToD would require Annie to be unsuccessfully plying her trade for 4 hours in an area where she was known without being seen by anyone willing to come forward to testify. An early ToD would put her closer to Eddowes condition. Would she have persisted so long before calling no joy and opting for rough sleeping somewhere like the Whitechapel graveyard?

Maybe she did sleep rough somewhere, which could explain why nobody saw her, etc. If so, it appears she then got up only to meet JtR who approaches her outside of Hanbury Street.

For question 2, for your "perhaps" were to be accurate, wouldn't her perceived temperature then have to flat line for the next 12 hours?

Who's, Eddowes? It's hard to say, subjective descriptions are not set measurements, and what one describes as 'cold' now, might be "not quite cold" or "some heat" etc on another occasion. Our sensory systems are not perfect, and the conditions can influence how we end up perceiving the same physical input. There's lots of examples in vision, like this one where the squares marked A and B are in fact exactly the same shade of grey. This is one of the reasons why subjective impressions are not considered valid measurements for this sort of thing. I seriously don't think it's a big deal that different people give different descriptions of their perceptions of different cases. How each would have responded had they swapped places, etc, is really just our subjective impressions about what their subjective impressions would be - and if that isn't going to be something highly influenced by our beliefs then nothing is, so I'm not sure we'll really make any progress that counts.

While we do have the benefit of more knowledge, the police did have the advantage of experience, even though the profession was in it's infancy. While Trevor is often derided on this forum, I believe sometimes unfairly, he was a professional and speaks from an entirely different perspective to the layman.

Anyway, I'm delighted that we can discuss our differences of opinion without the adversarial "yes it is - no it isn't" repetition that is pervading this thread at present.

Best regards, George

Yes, I agree completely.

- Jeff

[Fleetwood Mac]

You are fluent in English are you?

TOD is measured in hours, that is the common standard.

So, when Dr Phillips states: "at least two hours and probably more":

1) He is talking of hours in the build up to "more", hours is his criteria for assessment in his own statement.

2) Hours is the standard measure in Dr Phillips' profession. It follows that when he states "more" he is talking of more hours, in accordance with the standard measure, as opposed to more time which is meaningless and not in line with the standards used in his profession.

This has all been done ad nauseam over the course of several pages. For further details, see the previous pages on this thread.

At this juncture I'm much more interested in the claim: 'modern experts say Dr Phillips' assessment is unreliable'. This has been repeated ad nauseam and accepted by everyone on this thread. Sherlock Holmes is the latest to repeat it, and when asked for examples he has gone quiet (as per usual).

I'm pretty confident there will be experts in the field who have looked at Dr Phillips' assessment, although I can't find any examples online.

So, who were these experts and what exactly did they say in relation to Dr Phillips' estimate?