Does E-cigs or PG cause emphysema?

[mothakaf]

Does E-cigs or PG cause emphysema? Is there is a scientific evidence on that?

[surbitonPete]

If any of us thought that we wouldn't be doing it.

[Stylopora]

Cellular and Connective Tissue Changes in Alveolar Septal Walls in Emphysema
GORDANA VLAHOVIC, MICHAEL L. RUSSELL, ROBERT R. MERCER, and JAMES D. CRAPO

Duke University Medical Center, Durham, North Carolina; National Institute of Occupational Safety and Health, Morgantown, West Virginia; and National Jewish Medical and Research Center, Denver, Colorado

ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Emphysema is commonly defined as enlargement of airspaces distal to terminal bronchioles accompanied by destruction of alveolar walls, but without obvious fibrosis. Morphometric techniques were used to correlate changes in components of the alveolar septa surrounding enlarged airspaces in human emphysema with the mean linear intercept (Lm) of those airspaces. Alveolar and capillary surface density decreased with increased Lm, but the ratio of these surface densities to each other remained close to normal for mild to moderate increases in Lm. This suggests that the decreased gas exchange observed in emphysema is initiated by a total loss of septa and not by selective pathological changes of the microvasculature. Increases in septal wall thickness directly correlated with increases in Lm. For the mild to moderate emphysema lesions included in this study, an increase of 100% in Lm correlated with a 130% increase in the relative volume of the alveolar septal interstitium. Significant increases occurred in both elastin (0.14 to 0.56 µm3/µm2 basement membrane [BM]) and collagen (0.49 to 1.63 µm3/µm2 BM). The increase in elastin and collagen raises the possibility of a remodeling process in the connective matrix in alveolar walls. Whether or not the new connective tissue represents a disordered, nonfunctional regional response needs to be determined. Vlahovic G, Russell ML, Mercer RR, Crapo JD. Cellular and connective tissue changes in alveolar septal walls in emphysema.

INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
Emphysema is characterized by abnormal enlargement of the respiratory regions of the lung distal to terminal bronchioles, accompanied by destruction of the walls, and with loss of tissue per unit volume (1). There are two major forms of emphysema: panacinar and centroacinar emphysema (1). Panacinar emphysema involves airspace enlargement throughout the acinus and is thought to commonly arise as a result of a deficiency in synthesis or secretion of 1-proteinase inhibitor (1-PI). The most frequently observed form of emphysema, centroacinar emphysema, develops in the central portions of the acinus in close proximity to respiratory bronchioles and is predominantly associated with prolonged exposure to cigarette smoke (5). The pathogenesis of emphysema is still unknown: the most accepted hypothesis is based on an imbalance in proteases and antiproteases. That hypothesis is that tissue injury results from the actions of excess proteolytic enzymes liberated from inflammatory cells such as neutrophils and monocytes (8).

The role of cigarette smoking in the formation of emphysema is partly explained by recruitment of polymorphonuclear leukocytes and monocytes in the lower respiratory tract either as a consequence of epithelial injury caused by smoke or as a response to chemicals in smoke (5). The targets of proteolytic enzymes and free radicals liberated from polymorphonuclear leukocytes and monocytes are collagen, elastin, proteoglycans, and 1-PI, respectively. Damage to the proteolytic enzyme inhibitor accelerates and augments development of emphysema.

Lung changes characterizing emphysema have been studied in both humans and animals via different methodologies (9, 11). Most of the findings in these studies suggest that connective tissue, especially elastin, is a major target of destruction in emphysema. The architectural rearrangement and loss of gas exchange surface caused by elastin degradation in emphysema is generally thought to be irreversible. Animal models of emphysema have been created by intratracheal administration of pancreatic elastase, and these models demonstrate synthesis of new elastin (12). The enhanced deposition of elastin in these models has been used as a basis to challenge the relevance of these animal models to human emphysema. It is not known how the development of emphysema in humans correlates with specific structural changes of the lung parenchyma, such as whether or not destruction of the vascular bed is an early event and whether or not early interstitial changes involve production or loss of either of the primary connective tissue elements, collagen and elastin.

The goal of this study was to investigate structural changes of the walls of enlarged airspaces occurring in areas of mild and moderate human emphysema to determine if enhanced deposition or degradation of connective tissue occurs. Morphometric techniques combining light and electron microscopy were used to quantify changes in interstitial elastin and collagen, interstitial inflammatory cells, endothelial cells, and alveolar epithelial cells in areas of mild to moderate emphysema. This comparison of structural changes in lung tissue from areas of airspace enlargement to the structure of normal lung demonstrates that specific connective tissue changes are part of the early pathological events in emphysema.

[Stylopora]

Many studies have demonstrated that degradation of elastin plays a key role in the initiation of emphysema (9, 25), although the loss of complete alveolar septal walls as a coordinated event raises questions as to whether or not elastin is the major target of destruction. The pathogenesis of emphysema involves a variety of events, including free radicals, activation of inflammation (polymorphonuclear and monocyte recruitment), and a variety of cellular derived mediators or cytokines all of which cumulatively lead to proteinase inhibitor inactivation, membrane lipid oxidation, and proteinase liberation (26). Focal lung injury, also a characteristic of emphysema, can be explained by the inhomogeneous distribution of cigarette smoke components (the major cause of acquired emphysema), antiproteinase inhibitors, antioxidants, and cellular derived oxidants in the lung (26).

The progressive loss of complete portions of the alveolar septa during the formation of emphysema suggests that gas exchange in the early stages of emphysema primarily decreases due to the loss of entire segments of alveolar septa, and not to a selective loss of the alveolar microvasculature. The ratio of capillary to alveolar surface in the septal tissue of mild and moderate emphysematous lesions remains normal. The volume of endothelial cells per unit alveolar surface in these areas compared with nearby healthy areas of human lung also remained unchanged. Thickening of the interstitium could lead to a partial diffusion block for gas exchange; however, the magnitude of the interstitial changes is not sufficient for the diffusion block to be functionally significant in comparison to the effects of the loss of entire alveolar septal segments.

In animal models of emphysema, repair has been shown to occur after acute injury to the lung (9, 11, 12). For instance, Mercer described gaps in elastin fibers of two hamster alveolar septal walls that occurred as a result of elastin destruction after exposure of the animals to a single dose of pancreatic elastase (12). He also illustrated enhanced deposition of elastin fibers during repair of the injured areas. This study supports the hypothesis that tissue repair and remodeling are a critical component of the process leading to emphysematous lesions. It is likely that following periods of intense elastolysis there are periods of repair in which elastin is remodeled perhaps in a disordered state and thus contributing to the loss of elastic recoil in the function of emphysematous lungs. Alternatively, phases of remission and repair, with fewer neutrophils present and macrophages predominating, could be the most common pathological expression of early phases of emphysema.

The results of the current study show a significant correlation between degree of emphysema and the thickness of interstitium in the remaining alveolar septal walls. Substantially enhanced numbers of neutrophils were not found in the tissues studied. The lack of neutrophils could be due to the fact that the tissue was obtained from patients who underwent surgery, and presurgical therapy or abstention from smoking could have reduced the frequency of acute inflammation. Two primary connective tissue components, elastin and collagen, were found to be increased in relative volume in areas of emphysema. An increase in elastin- and collagen-producing cells is also an important indicator of tissue modeling or repair (12, 27). In the current study, interstitial fibroblasts were consistently found in close proximity to areas of elastin (Figure 6) and collagen in the alveolar interstitium of diseased areas. Quite often the long processes of interstitial fibroblasts enveloped adjacent connective tissue elements. In vitro studies have demonstrated potential mechanisms for a direct repair process. Cultured pulmonary fibroblasts derived from neonatal rats demonstrate an increase in tropoelastin messenger RNA (mRNA) and elastin synthesis after the cells are exposed to elastase and elastase-solubilized extracellular matrix peptides. In cultured fibroblasts incubated with matrix peptides but not treated with elastase, a significant reduction of tropoelastin mRNA and elastin synthesis occurred (28). This suggests that elastin synthesis occurs at the sites where both elastase and injured extracellular matrix elements are present. On the basis of these studies, elastin synthesis would be expected to occur in the very areas where elastolysis initiates an emphysematous-like lesion. The current report of a significant correlation between enhanced elastin and collagen deposition and the local degree of emphysema illustrates this concept of remodeling, although possibly disordered, as a possible step in the creation of the emphysematous lesion.

The early stages of emphysema included in the present study were primarily characterized by the presence of macrophages rather than neutrophils. Mononuclear phagocytes usually accumulate in large numbers in the lung in response to cigarette smoking (1, 6). In this study they were found to be significantly increased in emphysematous lesions. It is believed that macrophages play a role in the pathogenesis of the alveolar septal injury that characterizes pulmonary emphysema, and that they may be important especially in the pathogenesis of chronic tissue destruction (10, 29). Their significance as a source of proteolytic enzymes and in the release of proteinase inhibitors is still questionable. It has been suggested that macrophages have elastolytic ability owing to liberation of the elastolytic enzymes metalloelastase and cathepsin L (28). Cathepsin L is significantly elevated and its mRNA highly expressed in alveolar marcrophages obtained from bronchalveolar lavage fluid from smokers compared with nonsmokers. This supports the concept that alveolar macrophages contribute to the proteolysis of elastin as part of the process of lung destruction associated with cigarette smoking.

This study shows that the walls of emphysematous lesions contain increased amounts of elastin and collagen which is consistent with either loss of interalveolar septal walls normally low in connective tissue or enhanced synthesis of elastin and collagen in emphysematous areasor that both processes are occurring. An increase in synthesis of elastin and collagen would suggest a repair process but does not necessarily indicate functionality. Whether newly synthesized connective tissue in emphysema undergoes the full process of maturation to normal connective tissue that is able to perform its functional role or represents a disordered nonfunctional regional response needs to be determined.

[taz3cat]

You can get emphysema from breathing desert sand during a sand storm. I know some older folks that never smoked in their lives and they have emphysema. You can't live forever, I don't care what people say. Take a chance enjoy life,

[Krakkan]

Bases on this article from 1942 PG can actually help kill harmful bacteria in the air wouldn't that be crazy to think our second hand vapors might actually be helpful? lol



THE BACTERICIDAL ACTION OF PROPYLENE GLYCOL VAPOR ON MICROORGANISMS SUSPENDED IN AIR. I

O. H. Robertson M.D.1, Edward Bigg M.D.1, Theodore T. Puck Ph.D.1, Benjamin F. Miller M.D.1, and With the Technical Assistance of Elizabeth A. Appell 1 From the Department of Medicine, the Douglas Smith Foundation for Medical Research, the Bartlett Memorial Fund, and the Zoller Memorial Dental Clinic of the University of Chicago, Chicago

It has been found that propylene glycol vapor dispersed into the air of an enclosed space produces a marked and rapid bactericidal effect on microorganisms introduced into such an atmosphere in droplet form. Concentrations of 1 gm. of propylene glycol vapor in two to four million cc. of air produced immediate and complete sterilization of air into which pneumococci, streptococci, staphylococci, H. influenzae, and other microorganisms as well as influenza virus had been sprayed. With lesser concentrations of propylene glycol, rapid and marked reduction in the number of air-borne bacteria occurred, but complete sterilization of the air required a certain interval of time. Pronounced effects on both pneumococci and hemolytic streptococci were observed when concentrations as low as 1 gm. of glycol to fifty million cc. of air were employed.
Numerous control tests showed that failure of the glycol-treated microorganisms to grow on the agar plates was due to actual death of the bacteria. The means by which propylene glycol vapor produces its effect on droplet-borne bacteria is discussed and data relating the bactericidal properties of propylene glycol in vitro to the lethal action of its vapor is presented.
Atmospheres containing propylene glycol vapor are invisible, odorless, and non-irritating. This glycol is essentially non-toxic when given orally and intravenously. Tests on possible deleterious effects of breathing propylene glycol containing atmospheres over long periods of time are being carried out.

[Stylopora]

THE BACTERICIDAL ACTION OF PROPYLENE GLYCOL VAPOR ON
MICROORGANISMS SUSPENDED IN AIR. I
BY O. H. ROBERTSON, M.D., EDWARD BIGG, M.D., THEODORE T. PUCK, PH.D.,
AND BENJAMIN F. MILLER, M.D.
W~TH THE TEC~NIC~ ASSZST~rCE O~ Er.rz~r~ A. APPELL
(From the Department of Medicine, the Douglas Smith Foundation for Medical Research,
the Bartlett Memorial Fund, and the Zoller Memorial Dental Clinic of the
University of Chicago, Chicago)
PLATES 18 Am) 19
(Received for publication, February 27, 1942)
The idea of employing bactericidal mists as a method for controlling airborne
respiratory infection is not new, but until recently no one had succeeded
in producing by such means a sterile or relatively bacteria-free atmosphere
which could be tolerated by human beings. During the past few years new
methods of chemical air sterilization have been devised. These consist in the
dispersal of germicidal mists containing the effective chemical agents in such
small amounts as to be non-detectable or at least unobjectionable to persons in
the treated atmosphere. The compounds employed for this purpose are
believed to be non-toxic in the minute amounts present in the inspired air.
The initial report on this new approach to the control of air contamination was
made by Douglas, Hill, and Smith (1) in 1928. By means of a very fine spray of
electrolyzed sea water, containing NaOC1 with about 1 per cent available chlorine,
these workers were able to effect a marked or complete killing of Bacillus coli dispersed
in the air. The material appeared to be non-irritating in the concentration employedj
which was approximately 1 gin. of the chemical solution in two million cc. of air.
This paper apparently attracted little attention and it was not until 10 years later
that active development of the subject began.
In 1938 two publications appeared, one by Trillat (2) concerning the properties of
germicidal aerosols, and the other by Masterman (3) on air sterilization by spraying
or atomizing hypochlorite solutions. TriUat's earlier investigation, covering a period
of many years, dealt with problems of droplet infection and the various properties of
aerosols, and culminated in his discovery of the sterilizing properties of germicidal
aerosols. 1 Trillat found that certain germicidal agents which killed bacteria in the
test tube in dilutions not higher than 1:200, were capable of musing death of airborne
bacteria when dispersed in aerosol form in concentrations of 1 gm. of the chemical
substance in 5,000,000 cc. of air. He believed that this bactericidal activity was
due to direct interaction between the aerosol droplets and the bacterial particles.
Liquid aerosols consist of droplets 1 to 2~ in diameter, dispersed in air. An
erroneous use of the term aerosol has been introduced by commercial concerns who
have applied it as a trade name to certain wetting and detergent compounds.
593
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
594 BACTERICIDAL ACTION OF PROPYLENE GLYCOL VAPOR. I
Trillat tested a number of common bactericidal compounds and found that of these
only resorcinol and sodium hypochlorite were satisfactory. The other compounds
were either inactive as aerosols or too toxic, or proved to be irritating or unpleasant.
He states that resorcinol is the agent of choice as the odor of hypochlorite becomes
disagreeable after a time.
Masterman presents evidence for the germicidal effect of fine mists of sodium
hypochlorite. He found that 1 gin. of 1 per cent NaOC1 atomized in as large a volume
as forty million cc. of air would produce sterilization. Masterman concludes that
this marked bactericidal action is due to the HOC1 gas liberated from the mist and
not to an aerosol effect (3, 4).
During 1939 and 1940, two groups of workers in England, Pulvertaft and Walker
(5), and Two,t, Baker, Finn, and Powell (6), confirmed and extended Trillat's and
Masterman's observations. Pulvertaft and Walker tested various substances for
activity as germicidal aerosols and recommended a solution of resorcinol in glycerol
and water as satisfactory. These workers also found that NaOC1 was highly effective
and killed air-borne bacteria in a dilution of 1 gin. of 2 per cent NaOCI in six million
ce. of air. Their test microorganisms included pathogenic invaders of the respiratory
tract as well as non-pathogens. Twort (6) and associates carried out an extensive
investigation of the physical properties of aerosols, their droplet size and rate of
evaporation, and the effects of various germicidal agents on a number of different
microorganisms under a variety of conditions. Their most effective aerosol solution
"S 2'' contained l0 per cent hexylresorcinol and 0.05 per cent alkyl sulfate, "lorol,"
in alkaline propylene glycol. They reported bactericidal effects on certain nonpathogenic
microorganisms with extraordinarily small amounts of this material, e.g.
1 gin. to four billion cc. of air.
Andrewes and coworkers (7) published a brief confirmatory report on the use of
bactericidal mists for air sterilization, and in addition noted that a few viruses, including
that of influenza, are susceptible to the mist action as judged by the reduction
of their infectivity for mice.
Twort and Baker (8) have proposed another and quite different type of agent for
air sterilization, namely, certain kinds of smokes. Smokes from ignited cardboard
soaked with potassium nitrate or from incense were found to be highly effective.
They report that 1 gln. of the chemical substance dispersed in smoke form in 500
million cc. of air causes destruction of 95 per cent of air-suspended bacteria within
15 minutes.
Our earlier work in this field consisted in an investigation of the air-sterilizing
activity of certain bactericidal substances used as aerosols (9). We first employed
certain of the synthetic detergents studied by Miller and Baker (10),
since their activity in vitro gave promise of greater effectiveness as bactericidal,
aerosols than the compounds used by the French and English workers. Preliminary
experiments indicated only moderate effectiveness of aqueous solutions
of these detergents. However, when the water was replaced by a hygroscopic
vehicle such as propylene glycol, the aerosol activity was markedly
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
O. H. ROBERTSON, E. BIGG, T. T. PUCK, AND B. F. MILLER 595
increased. We found subsequently that propylene glycol and certain related
glycols themselves act as effective bactericidal aerosols. 2
As the work progressed it was found that propylene glycol in vapor form was
highly bactericidal, and that the marked and rapid germicidal action of propylene
glycol aerosol was due to vapor liberated from the small glycol droplets.
When pure vapor was employed, it was found to be more effective (12) than
an equal quantity of propylene glycol dispersed as an aerosol. For the purpose
of explicit exposition, the work will be presented in the sequence of its development.
Methods and Materials
Bacterial Suspensions.--Standardized bacterial suspensions were made by resuspending
the centrifugated sediment of an actively growing culture in nutrient broth
to a predetermined opacity corresponding to approximately one billion microorganisms
per cc. Water and other diluents were employed occasionally. The suspensions
were sprayed into the chamber with a Graeser atomizer (14). In some experiments
other atomizers producing coarser droplets were used.
Media for Recovering Bacteria from Air.--For Staphylococcus albus and most other
non-pathogens nutrient agar with 0.5 per cent added dextrose was used. Rabbit
blood agar was employed for recovering pneumococci and Streptococcus viridans;
sheep blood agar for hemolytic streptococci and staphylococci and chocolate agar
for ttemophilus influsnzae. It was found that chilling the agar plates before use
resulted in maximum recovery of bacteria from the air samples.
Test Chambers.--The chambers were made of glass and metal as shown and described
in Text-fig. 1. Wooden chambers were found to be somewhat less sarisfactory.
The air was gently agitated by means of a small fan rotating at a rate
sufficient to produce detectable air movement in all parts of the chambers. A metal
fan with 4 blades each 2½ inches long, run at 75 volts by means of a variable transformer
which gave a speed of 500 R.P.M., proved satisfactory for this purpose. The
fan was run for a period of only 5 minutes after the introduction of the bacterial spray.
Method of Sampling Air.--The number of viable bacteria recoverable from the
chamber air was determined by withdrawing a measured volume of air at a constant
2 Although the English workers used glycols and glycerin as vehicles, they apparently
ascribed little or no importance to these compounds beyond their usefulness as
hygroscopic solvents for the germicidal substances, resorcinol and hexylresorcinol.
The only reference to a possible independent action of propylene glycol is that made
by Twort and coworkers (6), who reported a single experiment in which solutions of
propylene glycol in alcohol exerted a very high degree of bactericidal activity when
dispersed in mist form. It seems probable, however, that the germicidal action of
this mixture was principally due to the alcohol, since alcohol vapor itself possesses
marked germicidal properties. Furthermore, Baker and Twort state in a recent
paper (13) that the presence of propylene glycol in their S 2 mixture contributes little
or nothing to its bactericidal activity.
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
~6 BACTERICIDAL ACTION OF PROPYLEN~ GLYCOL VAPOR. I
rate (2 liters in 2 minutes) through a glass funnel which was suspended directly abov(
the surface of an agar plate within a sealed glass jar (Text-fig. 1). This is a modifica
tion of the technique of air sampling described by Hollaender and DallaValle (15)
The rubber and glass connections are made as short as possible. The air vent foJ
equalizing internal and external pressure is covered with one or two layers of cantor
~cotton
Tower A
to suction pump I~ atomizer.l
....... " Jl
k- allar plate
T•xT-Fm. 1. Apparatus employed for determining bactericidal activity of glycol
vapors. A, 60 liter air-tight glass-walled chamber with one metal wall (opposite the
side with vents) fitted with a door dosing on rubber gaskets. (All sides, 15 inches
square.) The three upper orifices are actually on a horizontal line across the center
of the wall. B, Graeser atomizer connected with inlet by rubber tube. C, sampling
jar. G, air vent allowing water to return from bottle E to D.
flannel. When pathogenic microorganisms were being studied, the orifices of the
sampling jars were clamped and the jar containing the agar plate placed in the incubator.
Before opening, a considerable volume of air was drawn through the jar
and passed through a bead tower containing phenol (Text-fig. 1). Likewise, after
each sampling the air from the measuring bottle was expelled through phenol in the
bead tower. At the end of experiments with pathogens, the chambers were filled
with propylene glycol aerosol or vapor and this air drown through the suction bottles.
Canton flannel masks were worn as an added precaution.
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
O. H. ROBERTSON~ E. BIG% T. T. PUCK~ AND B. F. MILLER 597
Propylene GlycoL--The propylene glycol used in this study 3 was dear, colorless,
and odorless, and when fractionated at a pressure of 25 ram., more than 80 per cent
of the material was collected at a temperature of 99-101°C. (The boiling point
given in the literature is 100-101.8°C. at 24 ram.) Propylene glycol is relatively
non-volatile---its normal boiling point is 188°C., and at 200C. its vapor pressure is
0.18 ram. Hg. Thus a concentration of 0.73 rag. per liter in the vapor state represents
atmospheric saturation. No difference in behavior was detectable when the stock
propylene glycol or the redistilled material was used in the experiments described in
this paper.
Production of Propylene Glycol A erosol.--An effective aerosol was produced by means
of the Graeser atomizer which delivers a dry mist consisting of droplets averaging 2 to
3/t in diameter. The DeVilbis atomizer No. 180 is also satisfactory. 4
Production of Propylene Glycol Vapor.--The preparation of atmospheres containing
various concentrations of propylene glycol in the vapor state was accomplished in a
number of different ways. Placing the glycol in three Petri dishes on the floor of the
chambers, and allowing it to evaporate overnight (for 14 to 16 hours) results in a
vapor concentration of 0.35 rag. per liter, a quantity sufficient to produce almost
immediate killing of the various organisms tested in this study. More rapid and
complete saturation is effected by pouring propylene glycol, heated to 70--80°C.,
into three Petri dishes placed in the chamber. The chamber is sealed, and allowed
to cool to room temperature while a fan circulates the air inside. Still another
method used consisted in filling the chamber with air bubbled through propylene
glycol heated to 60°C. (in a thermostated water bath), which resulted in a glycol
concentration of 0.66 rag. per liter in the vapor state.
In order to achieve a more rapid and accurate means of filling the chambers with
any desired concentrations of propylene glycol, the apparatus shown in Text-fig. 2
was constructed. A 60 cycle synchronous motor of 1 R.1,.~r. speed drives forward a
screw (c) of 20 threads per inch by means of a worm drive (a) and gear (b), (ratio
7½ : 1) causing plate (d) to move forward and advance the plunger of a syringe. Propylene
glycol contained in the syringe is forced out at a constant rate on to a heater
which volatilizes the material. The heater consists of a 75 ohm, 10 watt "ohmite"
resistor, across which 30 volts are applied. The needle of the syringe rests on a wick
made up of a strip of cotton tape wound around the porcelain covering of the resistor.
The wick serves to disperse the liquid over the hot surface. An air stream whose
rate of flow is controlled by a pressure regulator and measured by a calibrated flow
meter, impinges on the heater and carries the volatilized glycol into the chamber.
The glass T tube (inner diameter 15 ram.) in which the heater is contained must be
as short as possible to obviate the possibility of condensation of the glycol on the
walls before it has been adequately mixed with the incoming air stream. Humidity
is controlled by a humidifier placed in the path of the air stream, as shown in Text-
We are indebted to the Carbon and Carbide Co. for supplying us with this highly
purified material. It should be pointed out that there are preparations of propylene
glycol on the market which contain impurities in toxic amounts.
4 We wish to express our appreciation to the DeVilbis Co. for making a special
experimental atomizer for us.

[Stylopora]

O. H. ROBERTSON, E. BIGG, T. T. PUCK, AND B. ~F. MILLER 599
fig. 2. The relative humidity can be varied either by changing the temperature or
the amount of the water in the humidifier. Relative humidity inside the chamber
was measured by means of wet and dry bulb thermometers, with the wet bulb placed
opposite the fan.
The concentration of propylene glycol vapor in the mixture issuing from the T tube
is readily calculated, since both the rate of delivery of liquid propylene glycol and
the rate of air flow through the vaporizer are known. Any desired concentration may
be secured either by changing the rate of flow of air through the flow meter, or by
inserting a syringe of a different diameter in the syringe holder. One cc. and 2 cc.
syringes were found to be most suitable.
The gas stream fed into the chamber is mixed completely with the air already present
by means of a rapidly rotating electric fan and then passes out through an escape
vent. A volume of gas four times that of the chamber is swept through for each
filling. (Calculation shows that after four air changes under these conditions the
propylene glycol content of the gas inside the chamber is 99.0 per cent of that in the
entering air stream.) The chamber is then sealed off. Before beginning the experiment
a sample of air is removed for analysis as an additional check on the glycol
concentration.
Method of Determining Concentration of Propylene Glycol Vapor in Air.--The determination
of the concentration of propylene glycol vapor was accomplished by
withdrawing a 2 liter sample of the air and bubbling it through 10 cc. of water with
the aid of a fairly porous, fritted glass gas-disperser. Complete absorption of the
propylene glycol is obtained at a sampling rate of 1/5 liter of air per minute.
The propylene glycol content of the resulting solution can be analyzed by the
method of Lehman and Newman (16), modified to accomodate the smaller concentrations
involved (17). In this reaction the propylene glycol is quantitatively oxidized
by a standard amount of periodic acid. The remainder of the oxidizing agent is
determined by reduction with a known quantity of sodium arsenite, and the excess
arsenite titrated with I2. The contents of the test tube are quantitatively washed into
an Erlenmeyer flask. 1.0 cc. of M/10 periodic acid ~ is added, and the sample is placed
in an ice box for 15 minutes. At the end of that time, 5 cc. of 7 per cent NaHCO3 is
added, then 2.5 cc. of N/10 Na~AsOz, followed by 0.4 cc. of freshly prepared 20 per
cent KI. The solution is allowed to stand for 15 minutes at room temperature,
after which 1 cc. of 1 per cent starch is added. The solution is then titrated with
0.01 N I2 to the end point marked by the appearance of the blue color of the starch
iodine complex. A blank is run through the same procedure, and the number of
milliliters of I~ solution used in the blank is subtracted from that required by the
sample. One cc. of 0.01 N I2 solution is equivalent to 0.38 mg. of propylene glycol.
Samples containing known amounts of propylene glycol varying from 0.3 to 1.0 mg.
gave results accurate to within 0.04 mg., when analyzed by this method.
5 The M/10 periodic acid is prepared by dissolving 5.35 gm. of sodium periodate
(NaIO4) in 75 cc. of N/1 sulfuric acid, and diluting to 250 cc. in a volumetric flask.
The resulting solution is stable indefinitely if stored in an ice box between runs.
Directions for preparation of the standard sodium arsenate and iodine solutions are
available in any standard textbook such as Pierce, W. C., and Haenisch, E. L., Quantitative
analysis, New York, John Wiley & Sons, 1937.
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
600 BACTERICIDALA CTION O]~ PROPYLENE GLYCOL VAPOR. I
EXPERIMENTAL
Reproduceability of Bacterial Counts in Air Samples.--Agar plates exposed
to the flow of bacteria-containing air by means of the technique outlined above
yielded a remarkably

Two different standard suspensions of staphylococcus were made up from the same
culture on each of two occasions. The dilutions of the standard suspensions and the
time of spraying are indicated in the first column of the table. The same atomizer
filled to the same level and operated at 500 mm. air pressure was used in all the tests.
Air samples (of 2 liters each) were withdrawn at exactly the same intervals following
termination of the bacterial spray. The first sample (labeled immediately) was
started 15 seconds afterwards and the second one at 5 minutes. Mter two samples
had been taken, the chamber was cleaned out and the next test run under identical
conditions.
It will be noted in the table that suspensions 1 and 2 yielded approximately
the same number of colonies on the plates while the difference in sampling
yields between suspensions 3 and 4 indicates that we were less successful in
making up suspensions of equal density.
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
O. H. ROBERTSON, E. BIGG, T. T. PUCK, AND B. F. MILLER 601
There is a progressive and quite constant diminution in the number of
colonies recovered from successive air samples during the period of an hour, as
illustrated in Plate 18, Figs. 1 to 4. This phenomenon may be ascribed to
settling of the bacterial droplets, inelastic collisions with the chamber walls,
and natural mortality of the bacteria. This necessitated the use of two identical
chambers for each experiment, one for the test, the other as a control.
Propylene Glycol Aerosol
Tests on the bactericidal activity of propylene glycol aerosol were performed
as follows: usually the bacteria were sprayed into the chamber first, and an
air sample obtained, following which a weighed quantity of the aerosol was
introduced. In order to control any possible effect of the germicidal aerosol
other than its direct bactericidal action, distilled water was sprayed into the
control chamber in each experiment. It was found that a concentration of
TABLE II
Effect of Propylene Glycol Aerosol on Staphylococcus albus
Aerosol
concentration
1:2,000,000
Time intervals of air samples
Immediately after bacterial spray
Immediately after H20 in control and aerosol
in test
15 min. later
30 rain. later
No. of colonies on plates from
Control
chamber
2140
1940
510
350
Test chamber
2124
3
1 gin. of propylene glycol aerosol in two million cc. of air effected complete
sterilization of an atmosphere into which as many as 500,000 bacteria per liter
of air had been sprayed. Furthermore, this action occurred with surprising
rapidity. Air samples taken within a few seconds after the introduction of
the aerosol yielded sterile plates while similar plates from the control chamber
showed many hundreds or thousands of colonies depending on the amount of
bacterial suspension used. A protocol of an experiment with Staphylococcus
albus is recorded in Table II and photographs of another experiment are shown
in Plate 18, Figs. 1 to 8.
A number of other microorganisms were found to be similarly susceptible to
the action of this aerosol. Among those tested were pathogenic invaders of
the respiratory tract, i.e., pneumococcus Types I and III, hemolytic streptococci,
and hemolytic staphylococci, as well as organisms of lesser or no pathogenicity,
such as Streptococcus viridans, Bacillus coli, Micrococcus catarrhalis, and
Bacillus subtilis (vegetative form).
The order of introduction of the bacteria and the propylene glycol into the
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
602 BACTERICIDAL ACTION OF PROPYLEN'E GLYCOL VAPOR. I
chamber was found to have no effect on the result. Thus an equally rapid
sterilization was obtained when bacteria were introduced into a test chamber
already containing propylene glycol. Other glycols, such as ethylene and
trimethylene, acted about as effectively as propylene glycol. Glycerin, on the
other hand, exhibited only a slight killing effect.
Propylene Glycol Vapor
(a) Bactericidal Effect of Aerosol Not Explainable on Basis of Droplet I~eraction.--
If the disappearance of bacteria in air treated with a propylene glycol
mist represents a true bactericidal effect, and evidence presented later indicates
that such is the case, how can this action be explained? The means by which
bactericidal mists produce a lethal concentration of the active agent in the
immediate environment of the bacteria would seem to be limited to two possibilities;
(a) direct contact between germicidal aerosol droplets and bacterial
particles; (b) production of sufficient vapor or gas by evaporation from the
germicidal droplets to permit rapid and abundant collision of gas molecules
with the bacterial particles. Trillat, Pulvertaft and Walker, and Twort and
his associates, believe that germicidal mists exert their antibacterial action
exclusively as aerosols. They state that the substances employed by them are
ineffective in the gas phase. In fact, their investigations of different agents
or mixtures for use as bactericidal aerosols have been directed toward developing
a mist with a very slow rate of evaporation. Masterman, on the other
hand, considers that the activity of the germicidal mist he employed, namely
NaOC1, is due to the liberation of HOC1 gas.
Calculations of the maximum number of contacts possible between aerosol
and bacterial droplets (which Twort et al. have made (6) and which we have
repeated) indicate that it would take between 2 and 200 hours for sterilization
to occur if this were the mode of action of the germicidal aerosol. Since
complete sterilization of a heavily contaminated atmosphere has been found
to take place in as short a time as 5 minutes by the English workers, and within
a matter of seconds by ourselves, the rate of interaction between the bactericidal
agent and air-suspended bacteria must be of an entirely different order of
magnitude. Such rapid interaction could occur only if the germicidal substance
were present in the gas phase.
(b) Postulation of Vapor-Droplet Effect.--Granted that rapid interaction
between gas molecules and droplets does occur, would the resulting concentration
of propylene glycol in the bacterial droplets be sufficient to produce the
striking bactericidal effects observed? Experiments in vitro showed that
propylene glycol in common with other closely related glycols exhibited relatively
low germicidal action. Certain microorganisms grow well in broth
containing as much as 5 to 15 per cent of the different glycols tested (propylene,
ethylene, and trimethylene). The pneumococcus, for example, grows in
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
O. H. ROBERTSON, E. BIGG, T. T. PUCK, AND B. ~F. MILLER 603
broth containing up to but not more than 5 per cent propylene glycol. Staphylococcus
albus is not inhibited until a concentration of more than 10 per cent
of this material is reached. However, when these bacteria are suspended in
80 to 90 per cent propylene glycol, they are kiUed immediately. Since prowlene
glycol possesses a marked affinity for water, rapid absorption of the vapor
by aqueous droplets might be expected to occur. Indeed, calculations show
that with vapor concentrations even considerably below the saturation value
of propylene glycol, the number of collisions between gas molecules and droplets
containing bacteria is sufficient to produce almost instantly, a lethal concentration
(70 to 80 per cent) of propylene glycol in the droplets. Furthermore,
the observed rate of evaporation of droplets of a propylene glycol mist
is so rapid that a relatively high vapor concentration is liberated within a
second or two. s
TABLE III
Effect of Propylene Glycol Vapor on Streptococcus hemolyticus
Vapor
concentration
1:7,700,000
Time intervals of air samples
Immediately after bacterial spray
5 min. later
15 min. later
30 min. later
60 man. later
No. of colonies on plates from
cChoanmtrboelr Test chamber
370 36
360 5
312 0
250 0
96 0
(c) Demonstration of the Bactericidal Activity of Propylene Glycol Vapor.-
Tests carried out under conditions identical with those in which the aerosol
was employed, showed that propylene glycol vapor was not only highly bactericidal
but acted more effectively than did the aerosol of this substance. In
carrying out such experiments the chamber was filled with glycol vapor of
known concentrations by means of one of the several methods described, following
which the bacterial suspension was introduced. Concentrations of not
less than 1 gm. of propylene glycol in three or four million cc. of air resulted in
immediate and complete sterilization of the chamber air. This effect was
demonstrated with staphylococci, pneumococci, hemolytic streptococci, H.
influemae, and H. pertussis. The results of an experiment in which pneumococcus
Type I was employed as the test organism are exhibited in Plate
6 The size and rate of evaporation of the droplets was determined by means of the
Millikan oil-drop apparatus (18). These calculations together with a detailed discussion
of th e physicochemical interactions of propylene glycol gas molecules and fine
fluid droplets will be presented elsewhere.
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
604 BACTERICIDAL ACTION OF PROPYLENE GLYCOL VAPOR. I
19, Figs. 1 to 6. With diminishing concentrations of propylene glycol,
immediate and marked bactericidal activity was still obtained, although complete
sterilization of the air required increasing intervals of time. Table III
shows the lethal action of a vapor concentration of 1:7,700,000 (0.13 mg. per
liter) on Streptococcus hemolyticus. The effectiveness of low concentrations
of propylene glycol vapor was found to depend also on other variable factors,
such as numbers of microorganisms dispersed into the air, the number of bacterial
droplets, humidity of the atmosphere, state of bacterial suspension, etc.
A detailed presentation of this phase of the work will be given in a second
paper. Suffice it to say here that under optimum conditions pronounced
bactericidal action of propylene glycol vapor against certain of the respiratory
pathogens could be demonstrated in concentrations as small as 1 gin. of the
glycol in 50,000,000 cc. of air.
Propylene glycol vapor was also found to exert a lethal or at least an inactivating
effect on the virus of influenza. This was determined by tests in
which the presence of the glycol vapor in concentration of 1:3,000,000 was
shown to protect mice completely against infection with amounts of air-borne
influenza virus that produced death regularly in the control animals (19).
Tests with Other Glycols
Experiments with other glycols in vapor form revealed variations in their
bactericidal activity. Ethylene glycol, 2,3-butylene glycol, trimethylene
glycol, and a number of compounds of related chemical composition were
found to be highly effective. The relationship between chemical structure
and bactericidal properties will be discussed in detail in a later communication
dealing with the mechanism of the action of glycol vapors.
Evidence for the Bactericidal Action of Propylene Glycol
Other workers in this field have assumed that lack of growth on agar surfaces
exposed to bacteria-laden atmospheres treated with germicidal mists
represents actual death of the bacteria in the air. It seemed to us, however,
that further experimental evidence was necessary to exclude the possibility
that some factor or factors present in the air containing the germicidal mist or
vapor might either inhibit growth or prevent adherence of the bacteria on the
agar surface. In order to test for any growth-inhibiting effect due to condensation
of the glycol itself on the collecting plates, the following control was
performed: agar plates were exposed directly to air saturated with propylene
glycol vapor or to the glycol spray from an atomizer, then used in taking
samples from the control chamber. They yielded just as many colonies as
plates not treated with glycol. The possibility that the reaction between the
propylene glycol vapor and the bacterial droplets might somehow change the
Downloaded from jem.rupress.org on February 23, 2009

[Stylopora]

Published June 1, 1942
O. H. ROBERTSON~ v.. BIGG~ T. T. PUCK~ AND B. F. MTLI.ER 605
TzxT-FIG. 3. Bead tower for collection of bacteria from chamber air. 20 to 25 cc.
of broth diluted with equal parts of sterile water is placed in a 20 inch glass cylinder
half filled with 3 ram. glass beads. This tower differs from that shown in Text-fig. 1
in that there are two monel metal screens of 24 and 70 mesh inserted at A.
TEx'r-Fic. 4. Device for collecting air-borne bacteria on a glass slide.
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
606 BACTERICIDAL 'ACTION OF PROPYLENE GLYCOL VAPOR. I
state of suspension of the latter and prevent their adherence to the agar surface
was tested by employing another method of coUecting the vapor treated
bacteria. Air from the chamber was drawn slowly through 25 to 50 cc. of
diluted nutrient broth in a glass cylinder containing many small glass beads
(Text-figs. 1 and 3). Hated samples of this fluid, through which air from the
control chamber had been bubbled, yielded a large number of colonies, whereas
samples of fluid through which the glycol-containing air had been drawn were
sterile.
TABLE IV
Inoculation of Mice wlgh Pneumococcus Type I Exposed to Propylene Glycol Vapor
Test
chamber
Control
chamber
Material
introduced into
chamber
Propylene
glycol vapor.
1:3,000,000
followed by
pneumococcus
spray
Pneumococcus
spray
Air Samples
T/me taken
Immediately after
bacterial
spray
10 rain. later
30 min. later
Immediately after
bacteria]
spray
10 rain. later
]No. of pneumococcus
colonies
Method ~ On
of late culture P--I~
Plate ¢
Bead
tower
Plate
Plate i12
Bead
tower
30 rain. later Plate 484
In fluid from
bead tower
1 cc. -- 228
Total --
5700*
Mice inoculated
with icc. of fluid
from bead tower
10 mice. All
died of pneumococcus
infeclion
in 24 to 36
hrs.
The presence of kiUed bacteria in the glycol-treated air was demonstrated
by condensing the moisture of the air drawn from the chamber on a chilled
microscope slide, as shown in Text-fig. 4. When these preparations were
stained, the bacteria appeared normal. Cultures of the condensed fluid on
agar and in broth showed no growth.
We have also eliminated the possibility that microorganisms, although
rendered incapable of growth on artificial media, might retain their capacity
to reproduce in a suitable host. Experiments were conducted in which highly
virulent pneumococci Type I were sprayed into a chamber containing the
propylene glycol vapor. The air was then drawn through sterile broth in a
bead tower and 1 cc. quantities of this fluid were injected into mice. These
10 mice. All
remained well
* The 2 liter air sample was drawn throu h 25 cc. of 50 per cent broth-water mixture.
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
O. H. ROBERTSON, E. BIGG, T. T. PUCK, AND B. F. MILLER 607
animals survived. However, when the procedure was performed with air
drawn from the control chamber, all the mice died of pneumococcic infection.
The protocol of an experiment of this type is shown in Table IV.
Further evidence of the bactericidal action of propylene glycol vapor was
provided by cultures made from the floor and walls of the chamber at the end
of the experiment. Whereas cultures from the control chambers always yielded
large numbers of bacteria, those from the test or glycol-treated chamber were
uniformly sterile.
Additional Data and Comments
The question as to how propylene glycol produces its rapid and marked
bactericidal effect has not been elucidated. However, certain observed characteristics
of the glycol-treated microorganisms indicate something of the
general nature of the effect. Gram-positive bacteria killed by exposure to
propylene glycol, either in vapor or liquid form, retain their Gram-positiveness
as well as their morphological integrity. Glycol-killed pneumococci show
typical capsular swelling in the presence of specific antipneumococci rabbit
serum and retain their antigenic properties. Mice, vaccinated with pneumococci
killed by propylene glycol, were found to be just as resistant to the
injection of living microorganisms as were mice similarly immunized with
heat-killed pneumococci. While pneumococci suspended in propylene glycol
retain their Gram-positiveness for many weeks or months, removal from this
medium results in their becoming Gram-negative. Such microorganisms freed
from the glycol undergo gradual dissolution. Autolysis is hastened by the
presence of bile salts. The change from the Gram-positive to the Gram-negative
state can be brought about by simply adding an equal volume of water
or physiological salt solution to the glycol suspension. These findings indicate
that propylene glycol inhibits but does not destroy the autolytic enzyme system
of the pneumococcus cell. Whether or not the other enzyme systems of
the pneumococcus are affected remains to be determined.
We have not made comparative studies of the effectiveness of propylene
glycol and the several compounds and mixtures previously employed by other
workers, except in the case of Twort's "S ~''. Our observations with "S ~''
(10 per cent hexylresorcinol in alkaline propylene glycol) have been confined
to tests on Staphylococcus albus. These experiments showed that the addition
of hexylresorcinol to propylene glycol increased markedly the bactericidal
activity of the latter substance. The difference between the two agents both
in aerosol and vapor forms was slight as judged by the immediate sterilizing
effect, but became apparent after 5 to 15 minutes. An increased lethal action
might have been anticipated from the presence of hexylresorcinol since, in
vitro, this substance has been shown to be bactericidal in dilutions of between
1:2,000 and 1:20,000 (depending on the manner of testing) in contrast to
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
608 BACTERICIDAL ACTION OF PROPYLENE GLYCOL VAPOR. I
propylene glycol which requires a concentration of 1:1, i.e., equal parts of
bacterial suspension and glycol to produce a similar effect. Even higher
concentrations of propylene glycol were needed to kill certain non-pathogenic
microorganisms. Since rapidity of action of the air sterilizing agent would
seem to be of great importance in the application of such a method to the control
of air-borne droplet infection, and since the addition of hexylresorcinol
appears to contribute little to the immediate bactericidal effect of propylene
glycol, we feel that for this and other reasons the use of the latter substance
alone is to be preferred. However, it may be desirable to employ, for certain
purposes, highly bactericidal agents such as hexylresorcinol even though they
possess much greater potential toxicity than does propylene glycol.
Atmospheres containing propylene glycol vapor in concentrations up to the
saturation point (0.7 mg. of glycol per liter of air or 1:1,400,000) are invisible,
odorless, non-irritating, and tasteless, except in the highest concentrations
when a faintly sweetish taste is detectable. Tests on the toxicity of propylene
glycol administered by the usual routes, i.e., oral and intravenous, have shown
this substance to he essentially non-toxic (20). While it would seem probable
that propylene glycol taken into the body by way of the respiratory tract would
be equally innocuous, such an assumption is not justifiable in the absence of
direct experimental evidence. We have carried out tests on the effect of
maintaining rats in atmospheres of propylene glycol vapor for a year or more.
The results of these experiments will be presented in the second paper.
SUMMARY
It has been found that propylene glycol vapor dispersed into the air of an
enclosed space produces a marked and rapid bactericidal effect on microorganisms
introduced into such an atmosphere in droplet form. Concentrations
of 1 gm. of propylene glycol vapor in two to four million cc. of air produced
immediate and complete sterilization of air into which pneumococci, streptococci,
staphylococci, H. influemae, and other microorganisms as well as influenza
virus had been sprayed. With lesser concentrations of propylene
glycol, rapid and marked reduction in the number of air-borne bacteria occurred,
but complete sterilization of the air required a certain interval of time.
Pronounced effects on both pneumococci and hemolytic streptococci were
observed when concentrations as low as 1 gm. of glycol to fifty million cc. of
air were employed.
Numerous control tests showed that failure of the glycol-treated microorganisms
to grow on the agar plates was due to actual death of the bacteria.
The means by which propylene glycol vapor produces its effect on droplet-borne
bacteria is discussed and data relating the bactericidal properties of propylene
glycol in vitro to the lethal action of its vapor is presented.
Atmospheres containing propylene glycol vapor are invisible, odorless,
Downloaded from jem.rupress.org on February 23, 2009
Published June 1, 1942
O. H. ROBERTSON~ E. BIGG, T. T. PUCK~ AND B. F. ~ILLER 609
and non-irritating. This glycol is essentially non-toxic when given orally and
intravenously. Tests on possible deleterious effects of breathing propylene
glycol containing atmospheres over long periods of time are being carried out.
BIBLIOGRAPHY
1. Douglas, S. R., Hill, L., and Smith, W., J. Ind. Hyg., 1928, 10, 219.
2. Trillat, M. A., Bull. Acad. m~d., 1938, series 3, 119, 64.
3. Masterman, A. T., J. Ind. Hyg., 1938, 20~ 278.
4. Masterman, A. T., J. Hyg., Cambridge, Eng., 1941, 41, 1, 44.
5. Pulvertaft, R. J. V., and Walker, J. W., J. Ityg., Cambridge, Eng., 1939, 39~ 696.
6. Twort, C. C., Baker, A. H., Finn, S. R., and Powell, E. 0., J. IIyg;, Cambridge,
Eng., 1940, 40, 253.
7. Andrewes, C. H., Lancet, 1940, 2, 770.
8. Twort, C. C., and Baker, A. H., Lancet, 1940, 2, 587.
9. Robertson, O. H., Bigg, E., Miller, B. F., and Baker, Z., Science, 1941, 93, 213.
I0. Miller, B. F., and Baker, Z., Science, 1940, 91, 624.
11. Robertson, O. H., Bigg, E., Miller, B. F., Baker, Z., Puck, T. T., Tr. Assn. Am.
Physn., 1941.
12. Robertson, O. H., Bigg, E., Miller, B. F., Puck, T. T., and Baker, Z., in Moulton,
F. R., Aerobiology symposium, The American Association for the Advancement
of Science, 1942, No. 17, in press.
13. Baker, A. H., and Twort, C. C., J. Hyg., Cambridge, Eng., 1941, 41~ 117.
14. Graeser, J. B., and Rowe, A. H., Am. J. Dis. Child., 1936, 52, 92.
15. Hollaender, A., and DallaValle, J. M., Pub. Health Rep., U. S. P. H. S., 1939,
54, 1,574.
16. Lehman, A. J., and Newman, H. W., J. Pharmacol. and Exp. Therap., 1937,
60, 312.
17. Puck, T. T., Science, 1942, 95, 178.
18. Millikan, R. A., The electron, Chicago, The University of Chicago Press, 1924, 90.
19. Robertson, O. H., Loosli, C. G., Puck, T. T., Bigg, E., and Miller, B. F., Science,
1941, 94, 612.
20. Hanzlik, P. S., Newman, H. W., Van Winlke, W., Lehman, A. J., and Kennedy
N. K., Y. Pharmacol. and Exp. Therap., 1939, 67, 101.

[DaveyRoots]

whoa! this is bigtime info! thanks alot guys, really helpfull.