I? 
 
\ 117 
 )3 
 >py 1 
 
 ^HE COiMPARATlVE VIABILITY OF SeEDS, FUNGl 
 
 AND Bacteria When Subjected to 
 Various Chemical Agents 
 
 A THESIS 
 
 UHMITTED TO THE FACULTY OF THE DEFAHTMENT 
 
 OF LITERATURE, SCIENCE AND THE ARTS OF 
 
 TUK UNIVERSITY OF MICHIGAN FOH 
 
 THE DEGREE OF DOCTOK 
 
 OF PHILOSOPHY 
 
 RICHARD D?: ZEEUW 
 
 ANN ARBOR 
 1911 
 
,N<^ 
 
 6^y 
 
 In cxcli'',i;,'^G 
 
 APR 16 1912 
 
 Abdruck aus dem 
 Ceutralblatt fiir Bakteriologie, Parasitenkunde und lufektionskrankheiten. 
 
 II. Abteilung. 
 Herausgeg. von Prof. Dr. 0. Uhlworm in Berlin. — Verlag von Gustav Fischer in Jena. 
 
 Bd. 31. 1911. No. 1/4. 
 
f 
 
 Naehdruck verboten. 
 
 The comparative Viability of Seeds, Fungi and Bacteria when 
 subjected to various chemical Agents. 
 
 Richard de Zeeiiw. 
 With 1 Textfigure. 
 
 The following- work was undertaken because of the lack of conclusive 
 evidence that it is possible to obtain seeds for experimental purposes, free 
 from contaminating organisms. 
 
 All of the work was done in the botanical laboratory of the University 
 of Michigan, under the supervision of Professor N e w c o m b e , to whom 
 I wish to acknowledge my indebtedness. For valuable aid and helpful sug- 
 gestions on the mycological side of the problem, I am greatly indebted to 
 Prof. J. B. P 1 1 c k , also of this laboratory. To Dr. F. G. N o v y of the 
 medical department, I am indebted for some valuable suggestions on the 
 bacteriological side, also to his assistant, Mr. W. A. Perkins, for valuable 
 practical aid. 
 
 Historical. 
 
 Very little has been done to study the comparative viability of seeds, 
 fungi and bacteria when treated with different disinfecting agents. It is only 
 within recent years that any work having that end in view has been done. 
 The attempted object is to obtain good seedlings, free from bacteria and fungi, 
 
Tlic uoiujjarativ(> Viability of Seeds, Fungi and l>aeteria when snbjeeted ete. 5 
 
 for experimental purposes. The earlier efforts along the line of sterilization 
 were to sterilize the hands, articles of clothing, rooms, etc. For the hands, 
 mercuric chloride mainly was used. Some valuable suggestions as to its anti- 
 septic action may be obtained from the work of ¥ ii r b r i n g e r and F r (^ y - 
 h a n (1897) and of D a n i e 1 s o h n and Hess (1902). The work along 
 these lines is suggestive of the jiossibility of obtainiiig seedlings, free I'roni 
 bacteria and fungi, for exjjerimental })urj)()ses. 
 
 After sulphur was found to be, not valueless, but inadequate, for room 
 sterilization, attention was more and more directed to formaldehyde. It is 
 conceded quite generally that formaldehyde is valuable only as a surface 
 disinfectant. Special emphasis is laid on this point by N o v y and W a i t c 
 (1898) in their j)aper on „I)isinfection of Rooms." they found tliat dried or 
 covered infectious material was not necessarily killed even after twenty 
 hours exposure to the gas. M ii Her (1901) also came to the conclusion that 
 only exposed bacteria are killed; further, that these nmst be vegetating 
 forms. R u b n e r and F e e r e n b o o m (1899) found that horizontal 
 surfaces might be sterilized by means of formaldeliyde, while vertical sur- 
 faces were not in the least affected. This they ascribe to a condensation of 
 the formaldehyde gas, causing it to settle in disinfecting quantities on the 
 horizontal surfaces, while the vertical surfaces receive no such deposit. 
 
 The M'ork done on seeds with formaldehyde is, as a general rule, very 
 unsatisfactory. K e h 1 e r (1904) did some work in that direction, but he 
 found that his seeds were more sensitive to the action of formaldehde than 
 the spores of either bacteria or fungi, when its action went deeper than the 
 mere surface, as is absolutely essential if one is to obtain sterile seeds. The 
 work of AV e r n e r (1904) indicates that a stronger solution of formaldehyde 
 than seeds can bear is necessary to kill even exposed bacterial spores. 
 C h e s t e r and B r o w n (1905), studying the action of formaldehyde in 
 milk, found that it took a one-eighth ])ercent solution over five days to kill 
 Bacillus s u b t i 1 i s. B o s c (1896) found that it took five hours to 
 kill pathogenic germs on cloth well exposed to formaldehyde gas. Stern- 
 b e r g (1901) and P a r k (1905) lay greater emphasis than any of the others 
 on the chemical action of formaldehyde in disinfection. Morse (1907) 
 found that formaldeliyde will kill P h y t o p li t h o r a i n f e s t a n s , a 
 ])arasitic fungus, on seed potatoes, without injuring the potatoes. 
 
 The best results to date in sterilization were obtained by the use of mer- 
 curic chloride. K e h 1 e r (1904) found that copper sulphate killed seeds so 
 qnickly that it was valueless as a disinfecting agent. He, however, obtained 
 excellent results with mercuric chloride. M i y a j i m a (1897) found that 
 a 0,3 per cent solution of mercuric chloride killed the seeds of Z e a m ays, 
 P i s u m sativum and V i c i a fab a. The harm was done by over- 
 exposure, the seeds being exposed for six hours or more. C z a p e k (1896) 
 found that he could obtain sterile seeds of Z e a mays by polishing the dry 
 seeds with a stiff brush till no more particles came off, then cleaning them, 
 thoroughly with a brush, soap and warm, sterile, distilled water, and finally 
 dipping them for two or three minutes in a 1 per cent solution of mercuric 
 chloride, and then rinsing them once in sterile distilled water. This treatment, 
 he claims, was sufficient to kill the fungus hyphae growing over the aleurone 
 layer, but not into it. 
 
 Freeman (1904), working also on L o 1 i u m t e m u 1 e n t u m , 
 treated the seeds with a 1 per cent solution of mercuric chloride, but found 
 
Q R. i c h a r d d e Z e e u w , 
 
 only some of the seeds to be free from the fungus. S t e w a r d (1908) found 
 that the seeds of Z e a mays when subjected for one-half to three-quarters 
 of an hour to a 0,5 per cent solution of mercuric chloride were still sterile at 
 the end of fourteen or sixteen days. P a u 1 and K r o n i g (1896) found 
 that a two percent solution of mercuric chloride was unal)le in twenty-five 
 minutes to destroy the spores of Bacillus a n t h r a c i s in a suspension 
 dried on the surface of tare-garnets. B e h r i n g (1888) attributed tlie failure 
 of mercuric chloride in sterilization to the formation of mercuric albuminates. 
 Nelson (1907), treating potatoes for Oospora scabies, found that 
 a 1 percent solution of mercuric chloride killed the })arasite without 
 injuring the potatoes. Some interesting work on the inhil)ition of 
 bacteria was done by Paul (1901). He found that a 1 to 1,000 000 solution 
 of mercuric chloride kept them in check without killing them. G e ]) - 
 pert (1889) found that a suspension of Bacillus a n t h r a c i s 
 dried on silk threads was prevented from developing by treating for ten 
 minutes with a 0,1 percent solution of mercuric chloride. However, after 
 having been treated for a half-hour with a 0,1 percent solution the organisms 
 were caused to develop by ]necipitating the mercury with ammonium-sul- 
 phate. Eriksson (1905) attempts to explain the overwintering and con- 
 sequently the persistence of wheat-rust under treatment by what he calls 
 the Mycoplasm theory. He assumes that part of the protoplasm in the cells 
 belongs to the host but that part of it belongs to the fungus and is to all ap- 
 perance dormant. Tliis theorv ap])ears rather fanciful in the light of the 
 researches of S t o w a r d (1908), H a n n i g (1908), B o 1 1 e y and P r i t - 
 chard (1905) and others. 
 
 Technique. 
 
 The work laid in this study was to test the action of the following agents: 
 cleaning fluid^), mercuric chloride, hydrogen peroxide, potassium dicliroraate, 
 ammonium persulphate, bromine water (common), and formaldehyde gas, on 
 the seeds of L u p i n u s a 1 b u s , P i s u m sativum, T r i t i c u m 
 V u 1 g a r e , H o r d e u m v u 1 g a r e , Z e a mays and S i n a p i s 
 alba. 
 
 The work can be convently divided into two parts. The first part has 
 to do mainly with the seeds, the second mainly with the fungi and bacteria. 
 
 The first thing to be done is to determine the highest concentration and 
 the longest period of exposure in each case which will still permit one to ob- 
 tain a fair percentage of normal seedlings, leaving out of consideration for the 
 time being the effect of the disinfectant on the adhering fungi and bacteria. 
 A great deal of unnecessary and tedious work would be involved in deter- 
 mining the effect of the disinfectant on the fungi and bacteria adhering to 
 the seeds while determining the effect on the germination of the seeds. The 
 germination tests were carried on during the early fall when the laboratory 
 had a normal temperature day and night of 22" C, This temperature was 
 fairly constant. Lots of twenty-five seeds each were exposed for different 
 lengths of time to different concentrations of each of the disinfectants, and, 
 after thorough washing, were placed in a Geneva germinator kept at room 
 temperature. They were left for four or five days and were then compared 
 with the controls. These controls were treated as nearly as possible in a 
 
 ^) Cleaning fluid = H0SO4 sp. gr. 1.83 saturated with KgCrOj. 
 
The coin2»arat.ive Vialiility of Seeds, l^^iiigi and Bacteria wlien suiijectcd etc. 7 
 
 luaniicr similar to iiie experiiiieiilal soeds, t'xcei)t that 110 (lisiiilVctant was 
 used. 
 
 After tlie highest i)()iiit of both exposure ajid coiu'entratiou had been 
 determined for each kind of seeds in the different disinfectants, each kind 
 of seeds thus treated was then tested to detei'mine whetlier bacteria and 
 fungi could enchire tliis treatment as well as the seeds. For determining the 
 latter point, an a|)paratus was constructed, the idea for which was obtained 
 from Kehlers (1904) paper. It differs from Kehlers ai)paratus in that the 
 seeds, after they are once placed in the disinfectant, are not exposed to con- 
 tamination in any form till after a two weeks incubation. K e h 1 e r trans- 
 ferred liis seeds by nutans of sterile forceps from the vessel in which they 
 had been treated to the flask of culture medium. 
 
 My api)aratus (see te t-figure) is constructed as follows: Tiu're is a. large 
 flask, A, which contains distilled water and is connected with a three-arm glass 
 cock, B. Flask A also has a glass tube, a, with a flange worked on the end. This 
 tube projects into the flask through the rubber stopper. Over the flanged end 
 of the tube is securely fastened a cap of cotton, b, to filter the air as it enters 
 the flask when the water is drawn ont of the flask. To the tube on the opposite 
 side of cock B, is attached a small flask C. This flask has also an upright 
 flanged tube, a', with a cotton filter I)'. To the third tube of cock A is attached 
 another small flask D. In addition to the flanged tube, a", with its cotton 
 filter b", it has a third tube, c, entering it. This tube is drawn to a point, dips 
 to the bottom of the flask, being bent so as to end in the angle between the 
 side wall and the bottom of the flask. Tube c is connected with cistern E, 
 which in turn is connected with an aspirator. Flasks A, C and D are supported 
 on a low ringstand. 
 
 The apparatus is operated as follows: Flask A is filled with distilled water; 
 into flask C is put 25 com of agar medium; and E is disconnected alter the 
 rubber tube between D and E has been closed by means of a clamp. The 
 whole apparatus (except the cistern) is then placed in the autoclave and 
 sterilized. After autoclaving, flask C, while still attached to the ringstand, 
 is introduced through a slot into a large cardboard cylinder. 
 
8 II i c li a r d d e Z e e ii \v , 
 
 Arotiiul the arm suppcrthig flask (' is packed sufficient cotton to close 
 the opening in the cylinder. E is again connected with the rest of the ap])aratus. 
 When the apparatus has cooled sufficiently, the rubber stopper is removed 
 from flask I), the seeds and the desired disinfectant arc quicldy introduced, 
 and the flask is again closed tightly. Flask D is then thoroughly shaken, 
 so that the disinfectant may come in intimate contact with the walls of the 
 flask, the tube and the stopper. This is to destroy any spores that may have 
 hapj)ened to enter when the flask was opened to admit the seeds and dis- 
 infectant. When the seeds have been in the disinfectant the recjuired length 
 of time, the disinfectant is drawn off into the cistern E, by means of tube c. 
 The last drops can readily be removed l)y tilting flask D. Water is then drawn 
 through cock B from flask A, into flask 1), upon the seeds. This is dom' by 
 exhausting the air in cistern E. To prevent the air from entering flask I) 
 during this process, a rubber finger-cot can be drawn over b, or the 
 finger may simply be pressed down on it. In that way the seeds can be was- 
 lied as frequently as desired. After washing them, some of the agar medium 
 is drawn from flask C by turning cock j^ so as to connect G and D. Agar 
 solidifies at 4-2" C. It is therefore kept at about 45" C until needed. That 
 temperature will not harm the fungi or bacteria that may adhere to the seeds. 
 Sufficient agar to cover the seeds is introduced. The rubber tubes between 
 D and C and between D and E are closed by means of clam])s. The rubber 
 tubes are cut beyond the clamps and flask D is ready be set aside to incubate. 
 
 Experimental. 
 
 In all of the following germination ex{)erinu'nts, all of the seeds were 
 carefully ins])ected so as to obtain only perfect s])ecimens. In all the germi- 
 nation experiments, each lot consisted of twenty-five seeds. The germination 
 })ercentage, under ordinary laboratory conditions, of the seeds in the follo- 
 wing experiments was as follows : L u p i n us a 1 b u s 96 per cent, P i s u m 
 s a t i V u m 96 per cent, T r i t i c u m v u 1 g a r e 96 per cent, H o r - 
 d e u m V u 1 g a r e 92 per cent, Z e a Mays 100 per cent, and S i n a ]) i s 
 a 1 b a 96 per cent. The first agent used in treating the seeds was cleaning 
 fluid (H2SO4 + K2Cr207). As far as I have been able to ascertain it has not 
 been employed for the ])urpose before. It occured to me that cleaning fluid 
 might be a very desirable agent to use since it penetrates no farther than 
 it thoroughly oxidizes. Thus, by timing its action, we can destroy as much 
 of the seed-coat as is safe under any given condition, and also the adhering 
 fungus and bacteria s})ores, even such fungus hyphae as may have pene- 
 trated part way into the seed-coat. From what I knew of the properties of 
 cleaning fluid, I had no idea that any seeds were so resistant to its action. 
 For instance, L u p i n u s a 1 b u s immersed in it for five hours and fifteen 
 minutes still gave 52 per cent of good seedlings. Up to about two hours 
 immersion L u p i n u s a 1 b u s gave, in practically every case, about 
 100 per cent of good seedlings. P i s u m sativum was also nu)re resistant 
 than was expected. After six minutes exposure to this fluid, 88 per cent 
 of good seedlings were obtained; and after forty-five minutes immersion, 
 the yield was still 76 per cent. T r i t i c u m v u 1 g a r e and H r d e u m 
 v u 1 g a r e were found to be rather sensitive to the action of cleaning fluid. 
 After seven minutes exposure, the former yielded only 28 per cent of good 
 seedlings; while after ten minutes exposure, the latter yielded 44 per cent 
 
The comparative Viability of Seeds, Fungi and Bacteria when subjected etc. 9 
 
 of good seedlings. Z e a mays was found to be more resistant than T r i t i - 
 cum or H r d e u m but not so resistant as L u p i n us or Pis u m. 
 After seventeen minutes exposure, 88 per cent of good seedlings ofZeamays 
 were obtained ; but from there on the percentage of germination fell rapidly. 
 Thus after twenty minutes exposure, only 16 per cent of good seedlings were 
 obtained; and, after twenty-five minutes exposure, only 4 per cent. S i n a p i s 
 alba was sliglitly more resistant than T r i t i c u m. After a ten minutes 
 exposure, 56 per cent of good seedlings were obtained. The foregoing tests 
 and the following w-ere made in a temperature of 20" to 22" C. 
 
 The second disinfectant employed w^as mercuric chloride. To prevent 
 the adhesion of air-bubbles, the seeds were first dipped in 70 per cent alcohol, 
 then thoroughly rinsed in water to remove the alcohol and finally placed 
 in the mercuric chloride. The concentration of mercuric chloride employed 
 varied from one gram of mercuric chloride in one thousand cc. of water to 
 one gram in fifteen cc. of water. When L u p i n us a 1 b u s w^as immersed 
 in a one to two hundred and fifty solution for fifty minutes, 100 per cent 
 of good seedlings were still obtained. Beyond that concentration and length 
 of immersion the percentage steadily decreased, until, at a concentration 
 of one to fifteen for one hour and fifteen minutes, it had dropped to 52 per cent. 
 For P i s u m sativum a one to five hundred solution only was used. 
 The yield of good seedlings, after a tliirty minute immersion, was 88 per cent; 
 and after a fifty minute immersion it was 64 per cent. Triticum vul- 
 gar e was found to be much more sensitive to mercuric chloride than either 
 L u p i n u s or P i s u m. After a thirty-minute immersion in a one to 
 five-hundred solution, only 52 per cent of good seedlings were obtained; 
 while, after an immersion of ten minutes, the yield was 92 per cent. H o r - 
 d e u m V u 1 g a r e w^as found to be extremely sensitive to the action of 
 mercuric chloride. Even after an immersion of only two minutes in a one 
 to one thousand solution, the yield of good seedlings w'as only 60 per cent. 
 Z e a mays was found to be more resistant to the action of mercuric chloride 
 than Triticum. After an immersion of thirty minutes in a one to five liundred 
 solution, the yield of good seedlings was 68 per cent; but in a one to two 
 hundred and fifty solution, after the same length of time, the yield was only 
 20 per cent. S i n a p i s alba, after an immersion of ten minutes, in a 
 one to one thousand solution yielded 76 per cent of good seedlings. When the 
 concontration was increased to one to five liundred and the time to thirty 
 minutes, the percentage dropped to eight. Before placing in the germinator, 
 the seeds were thoroughly washed in several changes of w^ater for about 
 an hour. 
 
 The third disinfectant employed was peroxide of hydrogen. The commercial 
 peroxide w\as used, and in no instance was it diluted. The length of time 
 the seeds were immersed in it varied from ten minutes to nine hours. L u - 
 p i n u s a 1 b u s , immersed for an hour, still yielded 100 per cent of good 
 seedlings; when immersed for nine hours, the yield had dropped to 60 per cent. 
 After an immersion of thirty minutes, P i s u m sativum yielded 
 100 per cent, but after one hour, the yield was only 52 per cent. Triticum 
 V u 1 g a r e , after being immersed for an hour, yielded 84 per cent of good 
 seedlings. H o r d e u m v u 1 g a r e was found to be very sensitive to the 
 action of the peroxide. After an immersion of only ten minutes, the yield 
 was 48 per cent, but when immersed for an hour the yield fell to 12 per cent. 
 After an immersion of nine hours, Z e a mays yielded 76 per cent of good 
 
20 RicharddeZeeuw, 
 
 seedlings. S i n a jj i s alba was found to be more sensitive, since, after 
 an hour's immersion, it yielded 64 per cent. 
 
 The fourth disinfectant employed was potassium dichromate. The 
 concentrations employed varied from N/2 to N/1000: and the length of 
 immersion varied from fifteen minutes to an hour. As in the case of the other 
 disinfectants, the yield of L u p i n us a 1 b u s was relatively high, since 
 the yield was still 84 per cent after an immersion of one hour in an N/2 solution. 
 The highest concentration found practicable, in the case of Pisum sati- 
 
 V u m was N/50. After being immersed in this for an hour, the yield was 
 72 per cent. The same concentration could be used for T r i t i c u m v u 1 - 
 g a r e. After a twenty-minute immersion, the yield was 88 per cent. Very 
 poor results were obtained with H o r d e u m v u 1 g a r e , since a twenty 
 minute immersion in a solution as dilute as N/1000 yielded only 40 per cent 
 of good seedlings. The concentration most useful in the case of Z e a Mays 
 was found to be N/10. The yield after an immersion of one hour, was still 
 56 per cent. N /25 was better adapted toSinapis alba than any other 
 concentration. After a twenty minute immersion, the yield of good seedlings 
 was 88 per cent. 
 
 Ammonium persulphate was next used. The highest concentration 
 used was one gram of the persulphate in three cc. of water the lowest con- 
 centration was one gram in one thousand cc. of water. The periods of immersion 
 varied from thirty minutes to four hours and a half. The longer periods 
 were used in the case of L u p i n u s a 1 b u s only. The following concen- 
 trations were found to be best adapted to the different seeds; one to three 
 for L u p i n u s a 1 b u s , one to two hundred and fifty for Pisum s a t i - 
 
 V u m , one to one hundred for T r i t i c u m v u 1 g a r e , one to fifty for 
 Z e a mays, and one to five hundred for S i n a p i s alba. When 
 L u p i n u s alba was immersed for four hours and a half in a one to three 
 solution, the yield of good seedlings was still 80 per cent. No higher concen- 
 tration could be obtained, since that was a saturated solution. The lowest 
 concentration, one to one thousand, was used for Hordeum vulgare only. 
 When Hordeum vulgare was immersed for only ten minutes, the 
 yield was only 28 per cent. 
 
 Since the halogens are good disinfectants, it was thought best to include 
 at least one of them in these experiments. For this purpose bromine water 
 was taken. Tlie plain bromine water, used as a reagent in the chemical labo- 
 ratory, was used. The bromine water was used full strength, as received from 
 the chemical laboratory; and in dilutions as low as one part of bromine water 
 to five thousand parts of water. The highest concentration was used for 
 Lupinus albus only, and the lowest for Hordeum vulgare 
 only. Lupinus albus immersed for an hour and forty minutes yielded 
 100 per cent of good seedlings. When immersed for six hours the yield was 
 44 per cent. Pisum s a t i v u m , immersed for ten minutes in a one 
 to fifty solution, gave a yield of 64 per cent. The behavior of T r i t i c u m 
 vulgare, at the different concentrations, was practically the same as 
 Pisum s a t i V u m , except that in the lower concentrations T r i t i c u m 
 gave a somewhat higher percentage of good seedlings Hordeum vul- 
 gare was so sensitive that it had to be thrown out. In a concentration 
 of one to five thousand in which it was immersed for only ten minutes, the 
 seeds were so much damaged that the yield of good seedlings was only 5 per cent. 
 
The comparative Viability of Seeds, Fungi and Bacteria when subjected etc. H 
 
 Concentrations of one to fifty, one to one hundred, and one to two hundred 
 and fifty were used for Z e a mays. The lenght of immersion varied from 
 ten to fifty minutes. The percentage yield varied from forty-four to seventy- 
 two. Although higher concentrations were tried for S i n a p i s alba, 
 one to five hundred was found to be the most suitable. After ten minutes 
 immersion in this solution, the percentage of good seedlings was ninety-two; 
 after fifty minutes immersion the percentage was sixty. 
 
 The last disinfectant used was formaldehyde gas. Forty per cent formal- 
 dehyde was put in an open vessel placed on a glass plate, on which the seeds 
 were also placed. The whole was then covered with a bell-jar, the edge of 
 which, where it came in contact with the glass plate, was given a coat of 
 vaseline so as to make the chamber air-tight. Both dry seeds and seeds soaked 
 for five minutes in water were used. The dry seeds of L u p i n u s a 1 b u s 
 had to be exposed for an hour and forty-five minutes before the germination 
 percentage dropped; and then it had dropped only to ninety-six. When the 
 soaked seeds had been exposed for forty-five minutes, the germination percen- 
 tage dropped to ninetytwo. Beyond these periods, the vitality of the seeds 
 steadily decreased, that of the soaked seeds becoming zero after seven hours, 
 while the dry seeds retained their vitality an hour or two longer. The rela- 
 tive effect of the gas on dry and wet seeds is clearly brought out in the case 
 of P i s u m sativum. Dry seeds, exposed for fifteen minutes yielded 
 64 per cent of good seedlings; while the soaked seeds yielded only 28 per cent. 
 To obtain a yield of 64 per cent in the case of the wet seeds, they could be 
 exposed for two minutes only. After a twenty-five minute exposure of the 
 wet seeds, only 16 per cent of good seedlings were obtained; while the 
 dry seeds, exposed for an hour, still yielded .36 per cent. Triticum vul- 
 gar e was found to be less sensitive to formaldehyde than P i s u m s a t i - 
 V u m. After an exposure of fifteen minutes, tlie dry seeds yielded 92 per cent 
 of good seedlings; while the soaked seeds still yielded 52 per cent. A ten- 
 minute exposure of the wet seeds, also gave 92 per cent. H o r d e u m v u 1 - 
 g a r e was found to be extremely sensitive to formaldehyde gas. A fifteen- 
 minute exposure of the dry seeds and a five minute exposure of the wet seeds 
 gave a germination percentage of only twelve. The dry seeds of Z e a mays 
 were considerably more resistant than Triticum ; the soaked seeds 
 only slightly so. Dry seeds, exposed for an hour, yielded 80 per cent of good 
 seedlings, while wet seeds, exposed for tliirty minutes, yielded only .36 per cent. 
 The dry seeds of S i n a p i s alba could be exposed three times as long 
 as tlie soaked seeds, and they still showed an equal vitality. Thus dry seeds 
 exposed for forty-five minutes, and soaked seeds, exposed for fifteen minutes, 
 both yielded 40 per cent of good seedlings. Dry seeds, exposed for thirty 
 minutes, yielded 68 per cent; while wet seeds, exposed for the same length 
 of time, yielded only 24 per cent. 
 
 In the first column of Table I are given the names of the seeds; in the 
 second column the disinfectants ; in the third column the concentrations best 
 suited for the different kinds of seeds; in the fourth column the length of 
 immersion of the seeds in the disinfectants w^hich would still give a fair per- 
 centage of good seedlings; in the fifth column the germination percentage; 
 and in the sixth column are given the results, as tested with the apparatus 
 (see plate), of the action of the different disinfectants on the fungi and bacteria 
 on the seeds. As can be seen by running over this column, only nine lots out 
 of forty were found to be sterile, each lot being treated differently. 
 
12 
 
 Richard de Zeeuw, 
 Table I. 
 
 
 
 
 
 % of 
 
 
 Seed 
 
 Disinfectant 
 
 Concen- 
 tration 
 
 Time 
 
 good 
 Seed- 
 lings 
 
 Result 
 
 Liipinus albus 
 
 Cleaning Fluid 
 
 Full Strength 
 
 3 hrs. 
 45 Min. 
 
 68 
 
 Sterile 
 
 Pisum sativum 
 
 5> 5? 
 
 55 55 
 
 10 min. 
 
 68 
 
 2 Sp. Fungi 
 
 Triticuiu viilgare 
 
 >> 55 
 
 55 55 
 
 3 „ 
 
 76 
 
 1 5, 
 
 Hordeum vulgare 
 
 5J 55 
 
 
 4 „ 
 
 76 
 
 1 55 
 
 Zea mays 
 
 
 55 ;5 
 
 15 „ 
 
 76 
 
 1 „ 5, 
 
 Sinapis alba 
 
 ,, ,, 
 
 55 55 
 
 8 55 
 
 72 
 
 Sterile 
 
 Lupinus albus 
 
 Mercuric chloride 
 
 1—15 
 
 1 hour 
 
 72 
 
 ,j 
 
 Pisum sativum 
 
 55 55 
 
 1—500 
 
 30 min. 
 
 72 
 
 Bacteria 
 
 Triticum vulgare 
 
 •5 55 
 
 1—500 
 
 15 ., 
 
 80 
 
 Sterile 
 
 Zea mays 
 
 55 55 
 
 1—500 
 
 20 „ 
 
 68 
 
 55 
 
 Sinapis alba 
 
 
 1—1000 
 
 10 „ 
 
 76 
 
 Bacteria 
 
 Lupinus albus 
 
 Hydrogen Peroxide 
 
 Full Strength 
 
 7 hours 
 
 76 
 
 Sterile 
 
 Pisum sativum 
 
 
 i^Lom.^ 
 
 45 minutes 
 
 76 
 
 Bacteria 
 
 Triticum vi^lgare 
 
 55 55 
 
 55 
 
 4 hours 
 
 80 
 
 Bacteria 
 1 Sp. Fungi 
 
 Zea mays 
 
 55 55 
 
 55 55 
 
 5 „ 
 
 84 
 
 Sterile 
 
 Sinapis alba 
 
 5' 55 
 
 55 55 
 
 45 minutes 
 
 64 
 
 ,, 
 
 Lupinus albus 
 
 Pot. Bichromate 
 
 N/2 
 
 1 hour 
 
 84 
 
 Bacteria 
 
 Pisum sativum 
 
 55 55 
 
 N/50 
 
 30 minutes 
 
 80 
 
 Bacteria 
 1 Sp. Fungi 
 
 Triticum vulgare 
 
 55 55 
 
 N/50 
 
 20 
 
 88 
 
 1 Sp. Fungi 
 
 Zea mays 
 
 55 55 
 
 N/10 
 
 30 
 
 72 
 
 1 55 
 
 Sinapis alba 
 
 5f 55 
 
 N/2o 
 
 20 
 
 96 
 
 Bacteria 
 1 Sp. Fungi 
 
 Lupinus albus 
 
 Ammonium Persulphate 
 
 1—3 
 
 4 hrs. 
 
 80 
 
 Bacteria 
 
 Pisum sativum 
 
 
 1—250 
 
 30 min. 
 
 80 
 
 ,, 
 
 Triticum vulgare 
 
 55 
 
 1—100 
 
 30 „ 
 
 72 
 
 Bacteria 
 1 Sp. Fungi 
 
 Zea mays 
 
 55 55 
 
 1—50 
 
 40 ., 
 
 68 
 
 Bacteria 
 1 Sp. Fungi 
 
 Sinapis alba 
 
 55 55 
 
 1—500 
 
 20 „ 
 
 80 
 
 Bacteria 
 
 Lupinus albus 
 
 Bromine water 
 
 Full Strength 
 
 1—1/2 hrs. 
 
 92 
 
 Sterile 
 
 Pisum sativum 
 
 55 55 
 
 1—100 
 
 15 minutes 
 
 68 
 
 Bacteria 
 
 Triticum vulgare 
 
 
 1—50 
 
 30 „ 
 
 72 
 
 1 Sp. Fungi 
 
 Zea mays 
 
 55 55 
 
 1—100 
 
 30 „ 
 
 72 
 
 1 ,5 
 
 Sinapis alba 
 
 (Seeds dry) 
 
 1—500 
 
 30 „ 
 
 80 
 
 1 „ 
 
 Lupinus albus 
 
 Formaldehyde gas 
 
 40% Formal 
 
 4 hrs. 
 
 72 
 
 1 55 55 
 
 Pisum sativum 
 
 •5 5' 
 
 55 
 
 15 minutes 
 
 64 
 
 Bacteria 
 
 Triticum vulgare 
 
 55 55 
 
 5. 
 
 30 „ 
 
 72 
 
 55 
 
 Zea mays 
 
 
 1 
 
 30 „ 
 
 64 
 
 1 Sp. Fungi 
 
 Sinapis alba 
 
 (Seeds wet) 
 
 55 55 
 
 30 „ 
 
 68 
 
 Bacteria 
 
 Lupinus albus 
 
 Formaldehyde gas 
 
 55 55 
 
 2 hours 
 
 72 
 
 ,, 
 
 Pisum sativum 
 
 55 55 
 
 55 55 
 
 1 minute 
 
 72 
 
 ,, 
 
 Triticum vulgare 
 
 55 55 
 
 55 55 
 
 10 minutes 
 
 92 
 
 J, 
 
 Zea mays 
 
 55 55 
 
 55 55 
 
 15 „ 
 
 50 
 
 1 Sp. Fungi 
 
 Sinapis alba 
 
 55 
 
 5? 55 
 
 5 55 
 
 80 
 
 Bacteria 
 
 I was not at all certain that the nine lots, which showed no contami- 
 nation in the foregoing experiments, would show the same results again, 
 if treated in a similar manner. For this purpose three lots of each were set 
 up, treated as before and allowed to incubate. The results are given in table 11. 
 
The comparative Viability of Seeds, Fungi and Bacteria when subjected etc. \^ 
 
 Four of the nine failed to show up sterile three times in succession, while five 
 remained sterile after incubation in all three tests. This means that appa- 
 rently only five lots out of the original forty could be depended upon as 
 being sterile when treated according to Table I. Of these five lots, three were 
 Lupinus alb us, and two were S i n a p i s alba. 
 
 Table II. 
 
 Seed 
 
 Disinfectant 
 
 Con- 
 centration 
 
 Sterility 
 
 Second 
 
 Lot 
 
 Third 
 Lot 
 
 Lupinus albus 
 Sinapis alba 
 Lupinus albus 
 Triticum vulgare 
 Zea mays 
 Lupinus albus 
 Zea mays 
 Sinapis alba 
 Lupinus albus 
 
 Cleaning 
 
 Fluid 
 Cleaning 
 
 Fluid 
 Mercuric 
 
 Chloride 
 Mercuric 
 
 Chloride 
 Mercuric 
 
 Chloride 
 Hydrogen 
 
 Peroxide 
 Hydrogen 
 
 Peroxide 
 Hydrogen 
 
 Peroxide 
 Bromine 
 
 water 
 
 Full Strength 
 
 1—15 
 
 1—500 
 
 1—500 
 
 Full Strength 
 
 3 hrs. 45 min. 
 8 minutes 
 1 hour 
 15 minutes 
 20 minutes 
 
 7 hours 
 
 5 hours 
 45 minutes 
 I — 1.1 hours 
 
 Sterile 
 
 Sterile 
 
 Bacteria 
 
 Sterile 
 
 Bacteria 
 Sterile 
 
 Sterile 
 
 Bacteria 
 Sterile 
 Fungus 
 Sterile 
 Fungus 
 Sterile 
 
 Since mercuric chloride is recognized as the most valuable desinfectant 
 we have, it was thought best to test for sterility three lots each of the seeds 
 used in these experiments. The length of time the seeds were immersed and 
 the concentrations are those determined upon in the germination tests with 
 mercuric chloride. By reference to Table I, the germination may be found. 
 The results of these three tests are given in columns three, four and five, 
 of Table III. None of these cultures showed growths of fungi or bacteria 
 three times in succession and none were sterile three times in succession. 
 The results in this case were rather disappointing, since they show that mer- 
 curic chloride cannot be depended upon to give sterile seeds in every case. 
 The more so, since the aim was to test kinds of seeds faii'ly representative 
 of all seeds likely to be used for laboratory purposes. 
 
 Table IIL 
 Mercuric Chloride. 
 
 Seed 
 
 First Trial Second Trial 1 Third Trial 
 
 Concentration 
 and time. 
 
 Lupinus albus 
 Pisum sati\'nm 
 Triticum vulgare 
 Zea mays 
 Sinapis alba 
 
 Sterile 
 
 Bacteria 
 
 Sterile 
 
 Sterile 
 
 Bacteria 
 
 Bacteria 
 Sterile 
 
 Bacteria 
 Sterile 
 Bacteria 
 
 Bacteria 
 Sterile 
 Sterile 
 Fungus 
 Sterile 
 
 1 — 15 for 1 hr. 
 1—500 „ 30 min. 
 1—500 „ 15 „ 
 1—500 „ 20 ,. 
 1—1000 „ 10 
 
 Since the decisive failure to obtain sterile seeds, shown in Tables I, II 
 and III, might reasonably be expected to make the technique appear open 
 
]^4 Richard deZeeuw, 
 
 to criticism, to say the least, it was thought advisable to set up a series of 
 control cultures by means of the apparatus. After the apparatus had been 
 sterilized in the autoclave, it was attached to the aspirator and air was drawn 
 throught he filter plugs for at least five minutes to test them and at the same 
 time to test the rubber connections. No seeds or disinfectant were placed 
 in D. Water was then drawn from flask A into flask D. After this a small 
 quantity of agar culture medium was drawn into flask D in the usual manner. 
 Flask D was then taken out and set aside to incubate, as in the experimental 
 cases. The idea was to test the apparatus by using it as nearly as possible 
 in the same way as when seeds and disinfectants were present. As a matter 
 of fact, I took less pains in the control experiments to see that the joints 
 were perfectly tight than I did when I tested seeds. The fact that flask D was 
 not opened during these control experiments cannot be held as an objection, 
 since any chance contamination, when the seeds were placed in flask D, 
 is reached by shaking up the disinfectant so as to thoroughly reach every 
 part that might have become contaminated. If the chance spores were able 
 to survive that, then it is manifestly impossible to kill those on the seeds, 
 and the point is proven either way. The only place the disinfectant, when 
 the seeds are treated, does not reach is the inside of the small glass tube con- 
 necting A and D. It was sterile when D was opened to admit the seeds and 
 disinfectant and, during the brief period of time that D remained open, the 
 mouth of the tube was directed downward, so no spores could drop in and 
 none could be drawn in, since there was no draft into it. Besides, if any conta- 
 mination could come from this source, in the case of the seeds, it would have 
 appeared in the controls as well. Yet none of the controls showed any conta- 
 mination. The conclusion seems irresistable that the contamination must 
 have come from the seeds. Twenty controls were used. 
 
 Discussion. 
 
 The following three points will be considered in the discussion of results: 
 
 I. Some causes of failure to obtain seeds free from fungi and bacteria 
 by disinfection methods. 
 
 II. Other methods that may be successful for obtaining seedlings free 
 from fungi and bacteria. 
 
 III. Some reasons why much of the work on seed-sterilization is open 
 to criticism. 
 
 I. 
 In view of the fact that out of six species of seed treated only two kinds, 
 Lupinus alb us and Pisum sativum, gave germinable seeds, 
 free from fungi and bacteria, it is only reasonable to look for some cause or 
 causes to explain this failure. The foregoing results are the more striking, 
 since, out of the seven disinfectants used, only three were effective in securing 
 seedlings free from bacteria and fungi, in the case of Lupinus a 1 b u s 
 and only two in the case of Sinapis alba. The cause of failure to obtain 
 the desired result is at least three-fold. First, the condition of the contami- 
 nating organisms or their environment may make it impossible to destroy 
 them, without destroying also the germ of the seed. Second ,the disinfec- 
 tants, in a given case, may act merely as antiseptics, producing only apparently 
 sterilized seeds. Third, the required concentration of the disinfectant and 
 the required length of immersion may differ from those outlined in the prece- 
 ding experiments. 
 
The comparative Viability of Seeds, Fungi and Bacteria when subjected etc. ]^5 
 
 Since conditions unfavorable for growth favor sponilation in bacteria, 
 it is probable that the bacteria are found on the seeds as spores. As is well 
 known, a bacterial spore is exceedingly more difficult to kill than a vegeta- 
 ting form. Add to this the protection the contaminating organisms enjoy 
 from small fissures in the seed-coat, loose epidermal cells, etc., and you have 
 a combination which aids the infecting organisms, but which gives little if 
 any protection to the extremely sensitive seed-embryo. N o v y and W a i t e's 
 ('98) work on room disinfection conclusively sliows that organisms, which 
 may be quite readily destroyed when freely exposed to the desinfectant, 
 will survive for hours, if treated in small masses, or if protected by small 
 amounts of foreign substances. There is no bacterium known that can be 
 immersed in a saturated solution of mercuric chloride for an hour and survive. 
 Yet I found that bacteria survived that treatment when found on seeds. 
 Protection of the organism can be the only explanation here. For this reason, 
 the Fur b ringer method might be useful in treating seeds. The value 
 of this method lies in the preliminary use of alcohol. The alcohol dissolves 
 and clears away from the seed-coat substances that interfere with the action 
 of the disinfectant which follows it. 
 
 Not only may the contaminating organisms be protected by fissures in 
 the seed-coat, debris, etc., but they may be found within the seed-coat. This 
 is especially the case with fungi. To explain the overwintering of the rust 
 on wheat, Eriksson formulated his Mycoplasm Theory. He claims 
 that the fungus is present in the host cells as naked proto})lasm, which could 
 be distinguished from the protoplasm of the host by careful staining. B o 1 1 e y 
 and Pritchard ('95) say that what Eriksson saw in the cells of 
 the seeds were the haustoria of the parasite. The Mycoplasm theory does 
 not seem necessary to account for the fact. B o 1 1 e y and P r i t c h a r d 
 ('05) have shown that the mycelium of wheat rust penetrates the seed-coat 
 and is thus carried over winter. H a n n i g ('08), working on L o 1 i u m 
 t e m u 1 e n t u m , found a mycelium inside the seed-coat, growing 
 over the aleurone layer. F r e e m a n ('04) also working on L o 1 i u m 
 t e m u 1 e n t u m , found that in some cases the mycelium had even pene- 
 trated the emliryo. W h e t z e 1 ('06), working on beans, found that a fungus 
 mycelium penetrated not only the pod but also the seed-coat and cotyledons. 
 My failure to remove fungi from seeds must, it seems to me, be referred to 
 a similar cause. It must be evident that when the fungus mycelium is wuthin 
 the seed-coat, as Hannig ('08) found in L o 1 i u m t e m u 1 e n t u m , or 
 with the embryo as W h e t z e 1 ('06) found in the bean, or as F r e e m a n 
 found in the L o 1 i u m t e m u 1 e n t u m , it becomes hopeless to expect 
 to kill it without at the same time killing the seed. 
 
 Of course the cases enumerated above were all instances of parasitic 
 fungi. But, if they show anything, they show that mere superficial disin- 
 fection will not avail in all cases. However, a saprophyte may become a 
 facultative parasite. It seems permissable to assume that many of the ordinary 
 fungi found on seeds in the laboratory behave as facultative parasites. That 
 they have penetrated the seed-coat to a greater or less depth cannot be 
 doubted. I found a striking confirmation of this fact in my work with cleaning 
 fluid. L u p i n u s a 1 b u s and S i n a p i s alba were found to be free 
 from infecting organisms of any kind. P i s u m sativum. T r i t i c u m 
 v u 1 g a r e , H o r d e u m v u 1 g a r e and Z e a mays were found to 
 be infected with fungi, after they had been treated with cleaning fluid till 
 
]^g Richard deZeeuw, 
 
 nearly all of the seed-coat or fruit-coat had been destroyed. P i s u m sati- 
 vum and Z e a mays even had two s])ecies of fungi growing on them 
 after treatment with cleaning fluid. On none of these six lots did bacteria 
 develope. These facts seem impossible of explanation, unless we assume 
 that the fungi had penetrated the seeds. 
 
 The failure to obtain seeds free from contamination may also be due 
 to the fact that the disinfectant acts as an antiseptic, which inhibis bacteria 
 but allows them to develope when it is removed. This brings up the question: 
 What is the difference between bactericidal and antiseptic action? Probably 
 the correct answer is that it is merely a nuitter of degree. Sternberg ('01) 
 makes the following distinction: "All disinfectants are also antiseptics, for 
 agents which destroy the vitality of the bacteria of putrefaction arrest the 
 putrefactive process: and these agents, in less amount than is required to 
 completely sterilize, arrest growth and thus act as antiseptics. But all anti- 
 septics are not germicides." Paul ('01) also says that inhibition, which 
 occurs when a solution of one to one million of mercuric chloride is used, 
 and sterilization are merely a matter of degree. Park (05)', in his work 
 on "Pathogenic Bacteria and Protozoa", recognized the following four degrees: 
 Attenuation, antisepsis, incomplete sterilization and disinfection. Anti- 
 sepsis has not been recognized at all in work on seed sterilization, and not 
 as much as it deserves in other work. It is very probable that in antisepsis 
 lies the key to the problem of obtaining seeds free from bacteria and fungi, 
 or rather, seeds on which the fungi and bacteria are inhibited from deve- 
 loping and multiplying. If by a dilute solution which will not harm the seeds, 
 we can prevent the organisms from developing, the result, as far as the bac- 
 teria and fungi are concerned, wUl be the same as if the seeds were actually 
 sterilized. Antisepsis is easily mistaken for disinfection, as Sternberg 
 ('01) points out: "One to ten thousand solution of mercuric chloride destroyed 
 the spores of B. a n t h r a c i s and B. s u b t i 1 i s in two hours. More 
 recent experiments show that failure to grow in culture solutions cannot 
 be accepted as evidence of the destruction of vitality in the cases of spores 
 exposed to the action of this agent, unless due precautions are taken to exclude 
 the restraining influence of the small amount of mercuric chloride." G e p p e r t 
 ('89), working on B. a n t h r a c i s , found that pieces of sUk thread, soaked 
 in a suspension of bacteria and then dried could be rendered apparently 
 sterile by leaving them for ten minutes in a one to one thousand solution 
 of mercuric chloride. Placed in l^ouillon and incubated for a sufficient length 
 of time, no colonies developed. But, if similar threads were used in a similar 
 manner, but treated with ammonium sulphide before being placed in the 
 bouillon, G e p p e r t found that abundant colonies developed. The ammo- 
 nium sulphide reacts chemically with the mercuric chloride breaking up that 
 powerful poison and forming the compounds ammonium chloride and mer- 
 curous sulphide. Several colonies were found to develope even in cases when 
 the pieces of thread had been treated with mercuric chloride for half an hour. 
 The objection that the silk threads afford protection to the bacteria might 
 reasonably be raised here. But G e p p e r t repeated his work with bacterial 
 suspensions dried on cover-glasses, with similar results. These facts show 
 that we are obliged to relegate many cases of so-called disinfection to anti- 
 sepsis. My own work with mercuric chloride on seeds also seemed to in- 
 dicate that the action was largely antiseptic. A marked difference was ob- 
 served, when the seeds were merely in two or three changes of sterile water 
 
The comparative Viability of Seeds, Fungi andjBacteiia when subjected etc. 17 
 
 and when they were washed for an hour in ten or twelve changes. Also a 
 wTak solution acting for a short time, but not so thoroughly washed off, was 
 more effective in producing antisepsis than a strong solution, acting for a 
 longer time and not so thoroughly washed off. At one time, when I had the 
 pleasure of discussing the question of antisepsis with Dr. N o v y , of the 
 bacteriological laboratory of this university, he laid special emphasis on the 
 size of the platinum loop with which inoculations were made from bacterial 
 suspensions, to which a definite amount of mercuric chloride had been added 
 as an antiseptic. He found, for instance, that enough mercuric chloride might 
 be transferred on a 2mm. loop, when a tube of bouillon was inoculated from 
 such a suspension, to inhibit the development of bacteria. A 1 mm. loop, 
 on the other hand, might carry over insufficient mercuric chloride to pro- 
 duce such results. Or again, if, after thoroughly mixing the bouillon in the 
 tube inoculated with a 2mm. loop, he inoculated a second tube of bouillon 
 with a loopful of the first he would get a growth of bacteria in the second 
 tube, while the first remained clear. 
 
 The last topic to be considered under the head of causes of failure to 
 obtain seeds free from contaminating organisms is the matter of the re- 
 quired concentration of the disinfectant and the length of time the seeds should 
 be immersed in it to produce the desked results. In other words: Is it, for 
 instance, better to immerse a seed for a short time in a strong, or, for a longer 
 time, in a weaker solution? There is always a possibility that this would 
 lead us nowhere. Still there is also the possibility that concentrations and 
 periods of immersion play a larger role in the relative effect on seeds and theii" 
 infecting organisms than we now suppose. To determine this point is beyond 
 the scope of the present paper. It would require a long series of experiments, 
 in which only one disinfectant is used and the results of which are constantly 
 checked by some such apparatus as was designed for the second series of ex- 
 periments in this paper. Some work done in this laboratory on imbibition 
 by seeds showed that about 80 percent of the water absorbed by the seed 
 was absorbed during the first five minutes. Thus it would seem that the more 
 rapid the action the better, provided it was stopped before it had reached 
 the embryo. According to different authors, the concentration of mercuric 
 chloride to be employed varied from 0.1 percent to 0.5 percent, and the re- 
 commended period of immersion from two or three minutes to half an hour. 
 No rule can be laid down, since one species of seed is so much more resistant 
 than another. Even different lots of the same kind of seed will vary so much 
 that the treatment which will kill one sample will not injure another sample. 
 For instance, K e li 1 e r (1904) immersed T r i t i c u m v u 1 g a r e in a 
 one to five hundred solution of mercuric chloride for half an hour and ob- 
 tained 99.5 percent of good seedlings. When I tried to duplicate his results, 
 I had only 52 percent of good seedlings. To complicate matters, the solution 
 does not keep the same concentration during the entire experiment, since 
 in some cases there is probably enough albumen on the seed-coats to combine 
 with most of the available mercury, forming an insoluble albuminate. All 
 of this goes to show how difficult it is, in work of this kind, to duplicate re- 
 sults. According to Novy and Waite (1895): „there is no chemic dis- 
 infectant which will invariably yield the same results regardless of the organism, 
 to be acted upon and the surroundings or environments of that organism''. 
 On comparing the results of other workers with each other and with my own, 
 I am convinced that in seed sterilization this holds preeminently true. An 
 
]^8 R i c h a r d d e Z e e u w , 
 
 examination of the work of P a m ni e 1 (1899) shows what part the seed- 
 coat plays in the variability of the results obtained in two sets of experiments. 
 Thickness, structure and composition, each adds something to the sum- 
 total of the results obtained. In this connection, see note on Browns paper 
 at the end of this paper, especially with reference to the extreme sensitiveness 
 of Hordeum noted in the preceding pages. It is not possible here to enter 
 into a full inquiry into and a complete discussion of the causes of the diffe- 
 rence observed in the relative sensitiveness of the different seeds. To do 
 that we should be obliged to devote too much space to the comparative histo- 
 logy of the seed-coats of the different seeds employed. 
 
 II. 
 
 There are three methods, other than disinfection, which should be dis- 
 cussed here. The first of these is excision of the embryo. This method has 
 been employed especially by men working on endospermic respu'ation, re- 
 spiration of the embryo, etc. It seems quite likely that sterile seedlings might 
 be obtained by this method. But the conditions must be just right. In the 
 first place, the embryo must be uninfected. When the fungus has penetrated 
 it, excision of the embryo is naturally useless. Further it requires the most 
 careful and delicate technique to obtain any results whatsoever by this me- 
 thod. How easily can an infecting germ be carried by the knife to the embryo, 
 when the parts covering it are removed ! If the seed-coat can be so thoroughly 
 sterilized that there is no danger of infecting the knife, there can be no neces- 
 sity for excising the embryo. 
 
 Another method is that adopted by C z a p e k (1896) for Z e a mays. 
 The dry grains were first polished by means of a dry, stiff brush, till no more 
 scales came off; then they were thoroughly scrubbed for several minutes 
 with warm water, soap and a brush; then they were rinsed in several changes 
 of warm sterile water; and finally they were immersed for two or three mi- 
 nutes in a one percent solution of mercuric chloride. The seeds were then 
 put in the germinator without first rinsing them. The main objection to 
 this method is its limited applicability. It is obviously impossible to treat 
 all seeds that way. Some seeds are too small, like Sinapis; or too light, like 
 many of the gramineae; while the surface configuration, like that of Tri- 
 t i c u m , precludes such a procedure. Another objection was the compara- 
 tively large amount of mercuric chloride that was allowed to remain on the 
 seeds, more than enough to cause antisepsis. No definite conclusion can 
 be reached without a series of experiments with that end in view, as to the 
 sterility or non-sterility of seeds thus treated. 
 
 A third method, employed by Harrison and Barlow (1 907), 
 may be called the selection method. About the only value I can see in this 
 method is the elimination of infected seeds. The method they followed is 
 as follows : One to three seeds were dropped in a test-tube containing about 
 3 cc. of boiling water. The tubes were immediately cooled and set aside. 
 They were tilted so that the seeds were only half covered with water. Those 
 tubes in which the water became cloudy, showing the presence of bacteria, 
 were rejected. The seedlings in the tubes, in which the liquid remained clear, 
 were then planted by means of sterile forceps in a sterile medium in E r 1 e n - 
 m e y e r flasks. These flasks were then set aside for four or five days. At 
 the end of that time, any cultures showing contamination were again eli- 
 minated. The writers do not say what percentage of the original seeds were 
 
The comparative Viability of Seeds, Fungi and Bacteria when subjected etc. 19 
 
 left. It certainly can not have been large. It is also doubtful that the hot 
 water acting for so short a time had any effect. 
 
 ITI. 
 
 In judging the value of previous work done on seed sterilization, it should 
 be borne in mind that practically all that has been done was merely inci- 
 dental. The author was, as a general rule, working on some problem for which 
 it was advantageous to have sterile seedlings. Thus we find in papers on 
 different pieces of work, tucked away here and there, a paragraph on seed 
 sterilization. In practically all instances there can be no doubt that sufficient 
 of the disinfectant remained to inhibit the development of such organisms 
 as might be present. This point has been more fully discussed above. But 
 what seems most remarkable is the lack of adequate proof that the seeds 
 were actually sterile. Thus Nelson (1907) treating seed-potatoes for 
 Oospora scabies, immersed them for an hour and a half in a one 
 to one thousand solution of mercuric chloride, without any subsequent was- 
 hing. They were allowed to dry and were then planted. Who shall say 
 how long a goodly amount of mercuric cliloride clung to the potatoes? The 
 same objection holds in regard to C z a p e k s (1896) work. Steward 
 (1908) claims that his seedlings were sterile at the end of fourteen or six- 
 teen days. The only proof he offers is the fact that he inoculated a tube of 
 bouillon with a little of the material scraped from a seedling with a platinum 
 needle. The prevailing tendency seems to be a too great willingness to assume 
 that the seedlings were sterile. K e h 1 e r (1904) obtained some striking 
 restdts. His technique, however, is not above criticism. It seems to me ex- 
 tremely doubtful that a rubber tube can be sterilized by boiling for half 
 an hour on three successive days the water in the flask to which it is attached. 
 The more so when the flask is closed by means of a cotton plug around the 
 tube entering it. This tube is withdrawn to above the water level during 
 the boiling and again pushed down after it has been flamed. Finally the 
 seeds are taken out of the receptacle and transferred to flasks of culture 
 medium through the air. The apparatus should have been construc- 
 ted in such a way as to prevent the seeds from coming in contact with the 
 air at any time until the incubation was completed. And yet, for all that, 
 K e h 1 e r claimed to have obtained sterile seeds in every case. K e h 1 e r 
 does not speak of thorough and prolonged washing. That probably explains 
 his results. Since K e h 1 e r mentions G e p p e r t ' s work Vv'ith ammonium 
 sulphide, it seems almost inexplicable that he did not check up his own results 
 by means of it. If he had done so and no organisms had developed, it would 
 be conclusive proof of the correctness of his conclusions. Now, it seems to 
 me, we are justified in doubting his conclusions. 
 
 Summary and Conclusions. 
 
 1. F r certain p li y s i o 1 o g i c a 1 experiments, seeds 
 free from bacteria and fungi are essential. Since 
 there was no convincing evidence at li a n d that a n y 
 of the methods used by others in disinfecting seeds, 
 were absolutely reliable, this work was undertaken 
 to supply, if possible, that evidence, or to prove 
 that the methods generally employed are i n a d e - 
 
20 
 
 Richard d e Zeeuw, 
 
 q u a t e to furnish seeds free from contaminating 
 organism s. 
 
 2. Lots of twenty-five each of the following spe- 
 cies of seeds were used in the foregoing experiments: 
 Lupinus albus, Pisum sativum, Triticum vulgare, 
 Horde um vulgare, Zea mays.andSinapis alba. Similar 
 lots in each case were treated for varying lengths 
 of time with each of the following disinfectants: 
 cleaning fluid, mercuric chloride, hydrogen per- 
 oxide, potassium dichromate, ammonium persul- 
 phate, bromine water, formaldehyde gas on dry 
 seeds and formaldehyde gas on seeds soaked in water 
 for five minutes. In the first series of experiments 
 the efforts were directed toward determining the 
 length of time each kind of seeds could be left in 
 each one of the disinfectants, and still yield 70 per- 
 cent or 80 percent of good seedlings. After this had 
 been determined, a second series of experiments was 
 used to determine the effect of the different dis- 
 infectants on the fungi, and bacteria, after the 
 seeds had been in the disinfectant for such a length 
 of time as would still permit a fair percentage of 
 seeds to germinate. 
 
 3. The results obtained are rather strikingly 
 opposed to those of other workers. When the action 
 of the disinfectant was stopped at the point where 
 still a fair percentage of the seeds were germinable, 
 the results showed quite uniformly that the con- 
 taminating organisms had not been destroyed, 
 except in a few instances noted below. Of the forty- 
 eight lots tested, only two lots of Sin apis alba and 
 three lots of Lupinus albus were free from bacteria 
 and fungi. The only disinfectants which removed 
 all contaminating organisms from these seeds were 
 cleaning fluid and peroxide of hydrogen. Bromine 
 water was also successful in the case of Lupinus 
 albus. The foregoing experiments to determine the 
 action of the disinfectants on fungi and bacteria 
 were set up with a specially constructed apparatus, 
 which prevented outside contamination. 
 
 4. In view of the results above, it was deemed 
 best to set up twenty control preparations. They 
 were set up in a manner similar to those above, ex- 
 cept that no seeds or disinfectant were used. All 
 of the control preparations were sterile at the end 
 of two weeks, proving t li a t the contamination, in 
 the case of the seeds, must have come from the seeds 
 themselves. 
 
 5. The foregoing work has convinced me that the 
 results of formerly published work are open to cri- 
 ticism in at least two respects: No adequate proof 
 
The comparative Viability of Seeds, Fungi and Bacteria when subjected etc. 21 
 
 is given that the seeds are really free from conta- 
 minating organisms and no means are employed to 
 remove the disinfectant so thoroughly that it can 
 no longer act as an antiseptic. In view of the fore- 
 going results, we are forced to conclude that t li e 
 majority of cases of so-called disinfection were 
 merely cases of antisepsis. 
 
 6. To antisepsis, and not to disinfection, we must 
 probably look for practical results. It makes no 
 difference in physiological experimentation whe- 
 ther a few dormant organisms cling to the seedlings 
 or not. What does the harm is their active growth 
 and multiplication. Absolute desinfection, which 
 seems out of the question at present, is not essen- 
 tial. 
 
 Note. 
 
 Since the completion of the foregoing work, which was not published 
 as soon as desired owing to unavoidable delays, several publications have 
 appeared bearing on the same topic. Two of these are especially worthy 
 of notice. Brow n^) has shown that the seed-coats ofHordeum vul- 
 gar e caerulescens are readily penetrated by mercuric chloride and some 
 other agents, while sulphuric acid and copper sulphate affect it less quickly. 
 He attributes this to a „selective action" of the seed-coat. I had noticed the 
 same thing, but did not have an explanation for it. This perhaps accounts 
 for the fact that I was unable to obtain seedlings from H o r d e u m , ex- 
 cept when I worked with cleaning fluid. 
 
 The second paper is by R o b i n s o n-). The purpose of this work was 
 to obtain ,,some definite knowledge of the effects produced by sterilization", 
 thus admitting at the outset that there is undoubtedly a residual effect of 
 sterilization that must be reckoned with. Thus the author takes practically 
 the same ground that I have taken in the preceding paper. 
 
 Both leguminous and non-leguminous seeds were used. Fifty seeds of 
 each of seven different kinds were used. The seeds were treated with the dis- 
 infectants, then rinsed several times in sterile, distilled water (a precaution 
 many do not use) and germinated on moist, sterile filter paper in sterile Petri 
 dishes. After the seeds had been in the Petri dishes for serveal days, platings 
 were made on beef agar from the seeds. 
 
 From the author's paper I infer that he did not subject his seeds to as 
 long and thorough washing as I did. That is the only way I can explain the 
 difference between his results and mine. For instance, he found that wheat 
 and corn were sterile after treating them for one hour with commercial hydro- 
 gen peroxide. In my work I found that they were not sterile even after treating 
 wheat four hours and corn five hours. The germination percentage was prac- 
 tically the same in his case and mine. The author himself calls attention to 
 the fact that enough disinfectant may adhere to the seeds to cause antisepsis. 
 Thus, some seeds of pea, wheat and radish were treated for thirty minutes 
 with a 0.5 percent solution of mercuric chloride, they were then washed 
 three times and the third wash water was used for plating Bacillus s u b - 
 
 1) Proceed, of the Rov. Soc. London. Vol. 81. Ser. B. p. 82. 
 
 2) U. S. Dept. of Ag/Bur. of Pit. Ind. Circ. No. 67. 
 
22 RicharddeZeeuw, 
 
 tills. At the end of the period of incubation the plates were sterile, sho- 
 wing that enough mercuric chloride remained on the seeds after two was- 
 hings to make the third wash water so toxic that it inhibited the growth of 
 Bacillus subtilis. Formaldehyde and hydrogen peroxide showed 
 the same results, unless used in a very weak solution and for a very short time. 
 The author also found that air-bubbles on or in the seeds interfered with 
 the action of the disinfectant. To overcome this difficulty he used a vacuum- 
 pump. The results are described as „good but not perfect". I found treating 
 them for a moment with 70 percent alcohol satisfactory. If the vacuum- 
 pump is used, the disinfectant is apt to penetrate too deeply to be readily 
 
 removed. 
 
 Bibliography. 
 A'b b a u. R o n d e H i , Das Atzsublimat und das Formaldehyd in der Desinfektions- 
 
 Vraxis. (Centralbl. f. Bakt. Abt. I. Orig. Bd. 33. 1903. p. 821 ; Ann. Rept. Bd. of Health 
 
 of Mass. vol. 33. 1905. p. 207.) 
 B e h r i n g , tjber Quecksilberalbuminat in eiweiBhaltigen JFliissigkeiten. (Centralbl. 
 
 f. Bakt. Bd. 3. 1888. p. 27.) 
 B o 1 1 e y , Einige Bemerkiingen iiber die symbiotische Mykoplasmatheorie bei dam 
 
 Getreiderost. (Centralbl. f. Bakt. Abt. II. Bd. 4. 1898. p. 855.) 
 B o 1 1 e y and Pritchard, Internal infection of the Wheat grain by rust. (Science. 
 
 N. Ser. Vol. 22. 1905. p. 343.) 
 B o s c , Essais de desinfection par le vapeur de formaldehyde. (Ann. de I'lnstit. Pasteur. 
 
 T. 10. 1896. p. 299.) 
 B r a a t z , tJber eine bisher unbeachtete Eigenschaft des Alkohols bei seiner Verwen- 
 
 dung zur Handereinigung. (Miinchner med. Wochenschr. Bd. 54. 1900. p. 1421.) 
 Burmester, Vergleichende Untersuchungen iiber den EinfluB der verschiedenen 
 
 Samenbeizmethoden auf die Keirafahigkeit gebeizten Saatgutes und iiber ihre pilz- 
 
 totende Wirkung. (Zeitschr. f. Pflanzenkrankh. Bd. 18 1908. p. 154.) 
 Chester and Brown, On the action of Formaldehyde in the preservation of milk. 
 
 (Centralbl. f. Bakt. Abt. II. Bd. 15. 1905. p. 629.) 
 C z a p e k Zur Lehre von den Wurzelausscheidungen. ( Jahrb. f. wiss. Bot. Bd. 29. 
 
 1896. p. 337.) 
 Danielsohn u. Hess, Alkohol und Sublamin als Handedesinfektionsmittel. 
 
 (Deutsche med. Wochenschr. Bd. 28. 1902. p. 1112.) 
 Eriksson, A general Review of the jirincipal Results of Swedish Research into 
 
 Grain Rust. (Bot. Gaz. Vol. 25. 1898. p. 26.) 
 — , The vegetative Life of some Uredineae. (Ann. of Bot. Vol. 19. 1905. p. 55.) 
 Freeman, The Seed-Fungus of Lolium temulentum. (PhUos. Trans. Roy, 
 
 Soc. of London. B. Vol. 196. 1904. p. 1.) 
 Fiirb ringer und F r e v h a n , Neue Untersuchungen iiber die Desinfektion der 
 
 Hande. (Deutsch. med. Wochenschr. Bd. 23. 1897. p7 81.) 
 G e p p e r t , Zur Lehre von den Antisepticis. (Berlin, klin. Wochenschr. Bd. 26. 1889. 
 
 p. 1182.) 
 — , t)ber desinfizierende Mittel und Methoden. (Berlin, klin. Wochenschr. Bd. 27. 1890. 
 
 p. 312.) 
 — , Die Desinfektionsfrage. (Deutsch. med. Wochenschr. Bd. 17. 1891. p. 797.) 
 G r a w i t z , Bemerkung zum Artikel von Mayer und W o 1 p e r t , tJber „Woh- 
 
 nungsdesinfektion durch Formaldehyd". (Hyg. Rundschau. Bd. 11. 1901. p. 395.) 
 H a n n i g , IJber pilzfreies Lolium temulentum. (Bot. Zeitg. Bd. 65. 1907. 
 
 p. 27.) 
 — , Die Bindung freien atmosphJirischen Stickstoffs durch pilzhaltiges Lolium te- 
 mulentum. (Ber. d. Deutsch. Bot. Gesellsch. Bd. 26. p. 238. 1908.) 
 Harrison and Barlow, The Nodule Organism of the Leguminosae — Its Iso- 
 lation, Cultivation, Identification and commercial Application. (Centralbl. f. Bakt. 
 
 Abt. IL Bd. 19. 1907. p. 264.) 
 K e h 1 e r , tJber die Sterilisation des Erdbodens und Pflanzensamen und iiber zwei ther- 
 
 moresistente Bakterien. [Diss.] Konigsberg i. Pr. 1904. 
 Hilgermann, Wasserstoffsuperoxyd als Reinigungs- und Desinfektionsmittel im 
 
 Friseurgewerbe. (Arch. f. Hyg. Bd. 14. 1892. p. 40.) 
 K e 1 h o f e r , tJber die Ausfiihrung und die Ergebnisse von Haftfestigkeitsversuchen 
 
 kupferhaltiger Bekampfungsmittel gegen die Peronospora. (Zeitschr. f. 
 
 Pflanzenkrankh. Bd. 17. 1907. p. 1.) 
 
The comparative Viability of Seeds, Fungi and Bacteria when subjected etc. 23 
 
 K r a e m e r , Dilute Sulphuric Acid as a Fungicide. (Proc. Anier. Phil. See. Vol. 45. 
 
 1906. p. 157.) 
 K r o n i g u. B 1 u m b e r g , Vergieichende Untersuchungen iiber den Wert der mecha- 
 
 nischen und Alkoholdesinfektion der Hande gegeniiber der Desinfektion mit Queck. 
 
 silbersalzen. (Miinchener med. Wochenschr. Bd. 47. 1900. p. 29.) 
 Mayer u. W o 1 p e r t , Zur RoUe der Lufttemperatur bei der Formaldehyddesinfek- 
 
 tion. (Hyg. Rundschau. Bd. 11. 1901. p. 396.) 
 — , Wohnungsdesinfektion durch Formaldehyd. (Hyg. Rundschau. Bd. 11. 1901. p. 153.) 
 M i y a j i m a , On the poisonous Action of Copper upon various Plants (Bot. Mag. 
 
 Tokvo. Vol. 11. 1897. p. 417.) 
 Morse , Potato Diseases in 1907. (Bull. Me. Agd. Exp. Stat. 149. 1907.) 
 M ii 1 1 e r , Vergieichende Untersuchungen iiber die desinfizierende Wirkung und die 
 
 raumliclie Verteilung des FormaldeJiyds bei dem Versprayungs- und Verdampfungs- 
 
 verfahren. (Centralbl. f. Bakt. Abt. I. Orig. Bd. 30. 1901. p. 454.) 
 Nelson, Some Potato Diseases. (Bull. Wyo. Exp. Stat. 71. 1907.) 
 N o V y a nd W a i t e , The Disinfection of Rooms. (Sep. Rep. to the Mich. State Bd. 
 
 of Health. 1898.) 
 P a m m e 1 , Anatomical Characters of the Seeds of Leguminosae, chiefly Genera of 
 
 Grays Manual. (Trans. Acad. Sci. of St. Louis. Vol. 9. 1899. p. 1.) 
 Park, Pathogenic Bacteria and Protozoa. (Lea & Febiger) New York. 1905. 
 Paul, Entwurf auf einheitliche Wertbestimmung chemischer Desinfektionsmittel, mit 
 
 besonderer Beriicksichtigung der neuen physikalisch-chemischen Theorien der Losungen 
 
 (Zeitschr. f. ang. Chemie. Bd. 14. 1901. p. 333.) 
 Paul u. K r o n i g , tjber das Verhalten der Bakterien zu chemischen Reagentien, 
 
 (Zeitschr. f. physik. Chem. Bd. 21. 1896. p. 414.) 
 P u r i e w i t s c h , Physiologische Untersucliungen iiber die Entleerung der Reserve 
 
 stoffbehalter. (Jahrb. f. wiss. Bot. Bd. 31. 1898. p. 1.) 
 R a h n , Die Empfindlichkeit der Faulnis- und Milchsaure-Bakterien gegen Gifte. (Cen 
 
 tralbl. f. Bakt. Abt. II. Bd. 14. 1905. p. 21.) 
 R e 1 1 g e r and E n d i c o 1 1 , The Use of Copper in the Purification of Water. (Eng, 
 
 N ws. Vol. 56. 1906. p. 425.) 
 R i d e a 1 , Disinfection and the Preservation of Food. New York (J. Wiley & Sons) 
 
 1903. 
 Rubner u. Peerenboom, Beitriige zur Theorie un ] Praxis der Formaldehyd 
 
 desinfektion. (Hyg. Rundschau. Bd. 9. 1899. p. 265 ) 
 Salmon. Further cultural Experiments with biologic Forms of the Erysiphaceae 
 
 (Ann. of Bot. Vol. 19. 1905. p. 127.) 
 — , On endophytic Adaptation shown l)y E r y s i p h e g r a m i n i s under cultm-al 
 
 Conditions. (Abstract of paper read before the Roy. Soc. of London. Apr. 6; 1905 
 
 Ann. of Bot. Vol. 19. 1905. p. 444.) 
 Sternberg, A Text-Book of Bacteriology. New York (W. Wood & Co.). 1901. 
 Steward, On endospermic Respiration in certain Seeds. (Ann. of Bot. Vol. 22. 1908, 
 
 p. 415) 
 Werner, Zur Kritik der Formaldehyddesinfektion. (Arch. f. Hvg. Bd. 50. 1904, 
 
 p. 305.) 
 W h e t z e 1 , Some Diseases of Beans. (Bull. Cornell Univ. Agr. Exp. Stat. 239. 1906. 
 
LIBRARY OF CONGRESS 
 
 DDDm337DliS