The airplane attack with gas occurred during a series of daily mustard gas attacks on Seicheprey from 21 - 25 March, when, according to Hanslian, a total of 1,325 77-mm and 105-mm yellow cross shells were fired in hour long attacks on the village. In the late afternoons or evenings of each of those days, amounts ranging between 160 and 350 gas shells were fired on the troop dugouts in Seicheprey and southeast of the village.71 It was part of the demonstrations taking place all along the Western Front, in conjunction with the first of the great German offensives in the spring of 1918, and was the only time a village in the 1st Division sector was gassed.
For almost a month previous to the German assault at Arras on 21 March,
the 1st Division had been on the alert as warnings were issued that the enemy was conducting "maneuvers of rupture" preparatory to a sudden offensive somewhere on the Western Front a "Harassing fire on trenches, communications, PCs, OPs, and artillery emplacements, especially with gas shells, can be expected for days preceding as well as Immediately before the assault."72 But the 1st Division, harassed and harassing in return almost daily, did not recognize the demonstration for the offensive.
The offensive began on the morning of 21 March when 64 specially trained German divisions, after a short artillery preparation with an unusually high proportion of gas shells (estimated at 200,000 to 250,000 blue and green cross shells) attacked the Third and Fifth British Armies between the Oise and Scarpe rivers on a 40-mile front from Arras to La Fere (see Map No. 1). By the end of the fifth day, the British had lost more than 150,000 men and the Germans had driven a 37-mile salient almost to Amiens, the capture of which would probably have separated the British and French armies. On 28 March, as the British armies rallied, the French made a furious counterattack at Montdidier, the point of deepest penetration, but could not retake the city. The German advance, however, had been brought to a halt.
Three days before, on 25 March, with the shattering of the allied lines on the Western Front apparently a matter of days, Pershing met with Petain
at Compiegne and put at his disposal all the American forces in France, to take their place in the battle line or to be used to relieve French divisions for the battle.73 The result was an order for the relief of the 1st Division in Lorraine. It was to proceed at once to Picardy, where the battle was then raging.74
Meanwhile, the 1st Division waited out the repercussions of the battle to the northwest. Capt. Louis S. Davis, who replaced Lieutenant Wright as Division Gas Officer on 20 March, did not report the gas attack on Seicheprey on 21 March, but said that 650 mustard gas shells fell in the village on the afternoon and evening of 22 March, that another attack was made with 200 shells on 23 March, and 250 more on 24 March - a total of 1,100 mustard gas shells. As of the 25th, there had been 19 men evacuated, all from the first attack. The attack on 24 March had resulted in no casualties "except a few eye cases" [sic].75 Whether any of the casualties occurred among the 800 replacements that were being equipped and trained in gas protection during the week of the gas attacks on Seicheprey, Davis did not say.
Although there is no evidence that Seicheprey was evacuated following
the first attack, Hanslian assumed that it must have been, in view of the few reported casualties, or else that much of the gas thereafter had fallen outside the town.
G - 3 reported the shelling of Seicheprey on three nights only in the period 21 - 25 March and indicated that the 20 gas casualties it reported had all resulted from the first bombardment with over 600 mustard gas shells on the evening of 21 March. G - 3 showed no casualties from the bombardments it reported on 24 and 25 March. G - 2 reported a single shelling of Seicheprey in the period, on 23 March. There was little agreement between G -2, G - 3, and the DGO on the shelling and no recognition of a pattern.76
Lt. Col. Hanson E. Ely, commander of the 28th Infantry, whose troops were in the Seicheprey area at the time, said nothing of the gas attack on 21 March but reported that approximately 450 mustard and chlorine shells had fallen in the area on the early evening of 22 March, causing 17 serious casualties. He was considerably exercised about the inability of the artillery to protect his men. When the gas shells began falling he called for retaliation fire but neither the gas nor the HE that had been fired had been sufficient to still the German batteries. His artillery liaison officer informed him "that it is very difficult for the artillery to get gas shells, and the amount they have on hand is not sufficient for a gas bombardment of this kind." Colonel Ely asked that more gas shells be obtained, to put the
division "on a par with the French and English for protection against such gas shell bombardments."
Queried by General Bullard, General Summerall pointed out the difficulty of silencing the German batteries. There were 56 of them against his 21 then in action. Even had sufficient gas shell been available, he said, up to noon on 23 March "the weather was favorable for emission of enemy gas; hence it was not favorable for our gas." In one 24-hour period, the enemy batteries had fired 720 shells [including gas shells?]. He had fired 1,390 shells, "including 600 retaliation." But on the afternoon and evening of 23 March, when meteorological conditions became favorable, said General Summerall, his artillery had put 400 gas shells on St. Baussant, Lahayville, and Camp de la Schlucht, a cluster of enemy dugouts east of Maizerais.77
The postwar summary on the use of special shells indicates that 171 gas shells were fired on a number of enemy targets on 21 March, 70 on 22 March, but with favorable weather, 70 155-mm and 559 75-mm gas shells were fired on 23 March and another 517 75-mm gas shells on 24 March.78 While these widely scattered "gas concentrations" with nonpersistent gases put down by the brigade may have harassed the enemy, they were not likely to
produce large numbers of casualties, particularly when any quantity of No. 4 (cyanogen chloride) shell was used. No. 4 gas tended to flash on burst and disintegrate.