10 BULLETIN 224, U. S. DEPARTMENT OF AGRICULTURE. 
a large part of the contents ran over the shells and collected around the eggs in the 
bottom of the pail, thereby making it impossible to prepare a clean product from 
leakers handled in this manner. 
TRAY METHOD OF HANDLING LEAKERS. 
To avoid these difficulties a tray method of handling leakers. was devised. The 
tray was made of galvanized iron and consisted of a drip pan upon which was placed a 
plate-with perforations for holding the leaking eggs. (Pl. XIII, fig. 2.) Thé openings 
were round and about 12 inches in diameter, or hexagonal with sides about six-tenths 
ofan inch in length. The plate was kept in position by means of solder in the corners 
of the drip pan, or by projections soldered on the sides about 1 inch from the bottom. 
The trays were | to 2 inches high, with perpendicular or beveled sides, and 1 foot 
square or 1 foot wide by 2 feet long, the former having a capacity of 3 dozen eggs, the 
latter 6 dozen. The smaller size was much more convenient because it took up less 
space in the candling room. 
After a leaking egg was candled it was placed, damaged end up, in one of the holes 
of the tray. When filled the tray was carried to the breaking room, where the eggs 
were broken and graded. The breaking and grading of these eggs was delegated to a 
few expert girls, because it was impossible to grade leakers as closely by the candle . 
or to break them in as cleanly a manner as cracked or sound eggs. The leaking eggs 
were opened with the thumb and the two first fingers of each hand, and in many 
instances without using a breaking knife. Precaution was taken to keep the leaking 
end of the egg down while the egg was being opened so that the contents would drop 
into the cup instead of running over the shell and wetting the fingers of the breaker. 
The same principles used in the grading of the regular breaking stock were used for 
leakers, except that the grading of soft eggs was done much more closely. 
During the first part of the season bacteriological and chemical tests were made of 
six small samples of leaking eggs collected in the candling room and opened and 
graded carefully in the breaking room. The results are given in Table 3. These 
results compared favorably with those obtained from contemporaneous samples of 
eggs broken from cracked and whole eggs, and warranted further investigation of 
leaking eggs to determine whether they could safely be conserved for food purposes. 
TABLE 3.—Evxperimental samples of leaking eggs, opened in the breaking room (D house, 
1912). 
Bacteria per gram Ammoniacal 
on plain agar in- ae miuroger fants 
cubated at— : method). Number 
Sample Date of collec- bacteria - A 
No. Source. tion. er eee OTOL ALD Moisture. pee 
Inlactose| wot Dry P 
20 C; 37 C. bile. basis basis. 
Per cent. | Per cent.| Per cent.| . 
4273 Dis May 'Sos eet. 2,100 1,300 10, 000 0.0015 0. 0053 71.91 13 
4274 1) 5 Meee do. 16, 500 4, 200 0 0013 0045 70. 83 13 
4275 Daal eee do. 45, 000 34, 500 10} 000i) secs. oe | semen eee ee eneeee 6 
4284 D=th| pare" do.. 37, 000 5, 000 00 0019 . 0067 W..53 12 
4286 D=Lrse ss do... 100, 000 62, 000 10 0013 . 0046 71.85 12 
Laboratory tests of three samples of leaking eggs broken in the candling room at F 
house during May showed a variation of from 1,600,000 to 25,000,000 per gram in the 
bacterial count, and of 10,000 to 100,000 per gram in the number of presumptive 
B. coli. (Table 4.) Sample No. 4370, representing 150 pounds of leakers, broken in 
the candling room of F house during the latter part of May, showed the high count 
of 25,000,000 organisms per gram, but a low amount of ammoniacal nitrogen, namely, 
0.0020 per cent on the fresh basis and of 0.0067 per cent on the dry basis. These 
results indicate that most of the eggs in the product were sound, but that there. 
were some highly contaminated eggs in the mass. Their presence was probably due 
to the impossibility of eliminating infected eggs when opening leakers in a candling 
room. Results of samples taken about a month later, but handled by the tray 
method, gave as shown 1n Table F—X (appendix) bacterial counts varying from 23,500 
to 1,700,000 per gram and the number of presumptive B. coli between 10 and 10,000 
per gram. These results indicated, therefore, that the minimum count of the samples 
of leaking eggs opened in the candling room was, approximately, the same as the 
maximum count of those opened in the breaking room; and that the presumptive 
number of B. col was also, in most cases, higher in eggs handled by the old method. 
